24:24
2025-01-27 13:13:59
2:06:12
2025-01-27 13:42:41
3:36:29
2025-01-28 09:08:14
4:33
2025-01-28 13:48:42
55:46
2025-01-28 14:06:51
2:02
2025-01-29 10:22:33
1:02:14
2025-01-29 10:25:14
2:10
2025-01-29 11:38:26
2:26
2025-01-29 12:03:00
1:23:37
2025-01-29 12:05:56
35:40
2025-01-29 15:01:26
1:40:43
2025-01-30 09:07:07
1:08:48
2025-01-30 11:20:20
1:10:50
2025-01-30 13:15:56
3:50:03
2025-01-31 07:20:07
Visit the Talend Big Data Integration y Ecosistema de Talend course recordings page
United Arab Emirates - Talend Big Data Integration
WEBVTT
-->
Thank you very much.
-->
Thank you very much.
-->
Thank you very much.
-->
Thank you very much.
-->
Thank you very much.
-->
Thank you very much.
-->
Thank you very much.
-->
So we're going to start with the example from yesterday,
-->
in which we're going to integrate the Big Data part with IA.
-->
How would we do that integration process?
-->
At first, we had already done a job that allowed us to
-->
connect to an IA and ask questions.
-->
The idea now is, let's say, to integrate some work done
-->
with Big Data, and then we can do an analysis or some
-->
projection using IA, in this case, the OpenAI.
-->
Next, we're going to continue with Hive, which is an engine.
-->
Basically, Hive is used a lot to form Data Warehouse or Data
-->
Lighthouse.
-->
And we continue with Spark, although with Spark I'm not going
-->
to give an example.
-->
I'm going to show you how it's done.
-->
Actually, it's very simple in the commercial version.
-->
It's nothing out of this world.
-->
And we end then with, oh, well,
-->
I'll give you an example of what the version part would be
-->
like with the proprietary tool.
-->
And we end with the part of the study case, evaluable.
-->
In the end, then, we do it in the last two hours.
-->
We're going to continue then, as I mentioned with the first
-->
example, which was the integration of Big Data with
-->
IA.
-->
For that, then, I'm going to connect now to the
-->
virtual machine, which I haven't connected yet.
-->
I'm going to see here.
-->
I think I started it.
-->
Yes, perfect.
-->
I'm going to share the virtual machine.
-->
It's here.
-->
Perfect.
-->
And the first thing we're going to do, then,
-->
is we're going to upload the Hadoop cluster that we have in
-->
Hyper-V.
-->
Then we go here to the direct access of Hyper-V,
-->
the administrator, and I just give it to start.
-->
Perfect.
-->
And I should leave it in the state that the virtual machine
-->
had yesterday.
-->
And then we're going to upload the one in Big Data,
-->
which we have here in the desktop.
-->
And that's it.
-->
Those are the two initial tools that we're going to see.
-->
We're done with this example.
-->
We're going to add a new tool called Hype,
-->
as I was saying.
-->
For that, I created a container.
-->
In this container, I have version 4 of Hype.
-->
I have to explain how the architecture is,
-->
how it's currently installed in each virtual machine.
-->
Let's go over here.
-->
Done.
-->
And there are some that are going up.
-->
There are some that already have it.
-->
Perfect.
-->
There are some that already have the one in Big Data installed.
-->
Perfect, then.
-->
Perfect.
-->
Here, then, it has gone up.
-->
And we're looking for the last job we did,
-->
which was to get the three products less sold.
-->
And we had made an initial filter by city and by gender.
-->
Particularly, I took as a city Yangon and as a male gender.
-->
We're going to look for that last job we did,
-->
which is here in the Big Data project.
-->
And it's called listing 10 products less sold.
-->
But in the end, with the filter, I had six records left.
-->
So we ended up only filtering the last three,
-->
that is, the three less sold.
-->
Perfect, then.
-->
Here I have that part.
-->
Done.
-->
I'm going to wait for some to be here.
-->
Well, this job, what it does,
-->
I had commented on a case with Netflix,
-->
and it's something very popular with Big Data.
-->
Normally, Big Data is used a lot for real-time
-->
or almost real-time in the case of Netflix,
-->
which analyzes everything you do.
-->
There may be a large amount of data that is going to be processed,
-->
but in the end, the result can be 10, 15, 20 records,
-->
as in this case.
-->
For this job,
-->
take into account that we take as an input from the Hadoop cluster,
-->
and remember that when we take as an input from the Hadoop cluster,
-->
what the tool does internally is that it converts that to MacReduce,
-->
and it would execute it against the cluster that you have configured.
-->
But this job has something in particular
-->
in that the output is sent directly to a file.
-->
From the fact that the file has few records,
-->
I think it lacks a bit of sense to send it to the Hadoop cluster,
-->
given the fact that they are 10 records
-->
and I can easily handle it as a traditional job.
-->
So that's why I sent it directly to a file
-->
and I didn't send it directly to the Hadoop cluster.
-->
Let's do the following then.
-->
Let's configure our PROM for this part.
-->
To configure this PROM, let's do the following.
-->
Go to the folder that we downloaded yesterday
-->
which is called Credentials OpenAI.
-->
To vary a bit and also so that it doesn't damage
-->
the PROM that we had with the previous job,
-->
we are going to create a file with a new PROM,
-->
so that we have the two PROMs.
-->
So there is a file that says prom.json.txt.
-->
Please take that file and take out a copy.
-->
Copy. If you want, paste it there.
-->
So here I am going to change the name.
-->
PROMPT.
-->
I am going to call it PROM less sold items.
-->
There I made a mistake of leaving it in TXT,
-->
but remember that the tool has no problem with that part.
-->
In the end, what it does is change the filter,
-->
but the important thing is the internal format,
-->
not so much the extension that it has,
-->
because the tool can read it.
-->
The idea of this is to create a new file
-->
so that it does not damage the previous job
-->
and also to see how we would do
-->
to change the initial configuration
-->
that we had of the metadata,
-->
because we are going to have a new file.
-->
For this, then, we are going to enter the new file.
-->
Well, I'm going to write a PROM.
-->
PROMs are not very good at these PROMs.
-->
I'm going to write a PROM.
-->
I don't know if you want to write one of your own
-->
or follow me in the writing of this PROM.
-->
Remember that by the format of the file,
-->
we are going to write the PROM in the double quotes.
-->
So I'm going to start this way.
-->
I have a company that sells products in the city
-->
of Yangon, I don't know what it's called,
-->
I don't know, it sounds like India.
-->
I have a company that sells products
-->
in the city of Yangon
-->
to men and women.
-->
I have a company that sells products
-->
in the city of Yangon to men and women.
-->
Currently, men consume little of the following products.
-->
And here is where the trick comes from,
-->
but this trick is going to be done with Java.
-->
That is, I am going to insert the data
-->
that the Big Data brings me
-->
just after that product.
-->
Give it two points.
-->
This is important, regardless of the PROM
-->
that you want to do,
-->
you are going to put the following,
-->
percentage S,
-->
percentage S,
-->
two points, percentage S.
-->
That percentage S, what it's going to do
-->
is that in execution time,
-->
I'm going to replace a string
-->
where that percentage is.
-->
Currently, men consume little
-->
of the following products.
-->
I want you to make me a sales plan
-->
taking the cultural and socio-economic part
-->
of the region.
-->
If you want, I'll send you the text
-->
that I put here.
-->
Well, I don't know if you want to do it.
-->
The important thing is the percentage S.
-->
That percentage S has nothing to do with talent,
-->
it's more about Java.
-->
There I sent you the PROM
-->
through the chat.
-->
Please remember to save it.
-->
And I repeat, the PROM,
-->
just in case you are writing it.
-->
I started this way, it is not the best PROM,
-->
but the idea is the exercise of how we integrate it.
-->
I have a company that sells products
-->
in the city of Yangon to men and women.
-->
Currently, men consume little
-->
of the following products.
-->
I want you to make me a sales plan
-->
taking the cultural and socio-economic part
-->
of the region.
-->
The percentage S is important.
-->
Perfect, ready then.
-->
Let's save it here.
-->
Let's go back to Mr. Talent Big Data.
-->
And look what we are going to do now.
-->
We are going to take the job.
-->
We open the job base from yesterday.
-->
We made the one from yesterday.
-->
It is called Job Connection API Open AI.
-->
And copy all the job.
-->
I think if you right click,
-->
it gives you to copy all the job.
-->
Right click on the sub job you have there.
-->
And you have to copy all the elements
-->
of that sub job.
-->
And you are going to paste it
-->
just
-->
like that.
-->
Yes, here it is.
-->
And you paste it in the bottom part.
-->
You paste it in the new one
-->
which is the Job List 10 less sold items.
-->
It could be a generic job
-->
given a file
-->
but I don't think
-->
I have time for that generic job.
-->
Try to place it well down here
-->
so that it doesn't get confused.
-->
I am going to close
-->
I am going to add a little more here
-->
and I am going to minimize
-->
here in a minus that I have here
-->
so that it is more expanded.
-->
Let me know then
-->
if there are some that are still ready.
-->
Perfect.
-->
First, I am going to modify the PROM.
-->
Why? Because the PROM
-->
now I have to add a different element
-->
now. So look what I am going to do here.
-->
The line
-->
that connects the PROM with the
-->
T-Map, you are going to erase it.
-->
The PROM connection line
-->
with the T-Map.
-->
But there is still a moment
-->
that there are some that are still organizing
-->
the moment to see.
-->
It seems that
-->
a moment has been left.
-->
It seems that
-->
there is a moving window that does not allow
-->
to move here. Give me a moment then
-->
and I enter interactive
-->
to see something.
-->
Ready then.
-->
Super weird.
-->
Oops.
-->
I am going to close then the engineer
-->
let's see because how weird
-->
that happens.
-->
A moment to see then.
-->
Wait a moment
-->
please the others.
-->
I do not know why I have the CPU so
-->
let's see here. Let's try to open
-->
again. I think it is the JOK
-->
I do not know.
-->
There was another JOK to see.
-->
This is weird.
-->
It was here.
-->
I think it was the tab.
-->
A moment please.
-->
Being honest I do not know what happened there.
-->
But let's try to open
-->
the other again.
-->
Being honest I do not know what happened there.
-->
Let's try to open the JOK
-->
a moment please.
-->
Let's try to open the JOK a moment please.
-->
Being honest I do not know what happened there.
-->
Being honest I do not know what happened there.
-->
I am going to copy this man
-->
and paste it.
-->
I click here and correct the error.
-->
I do not know why it happens there.
-->
Remember to click here because
-->
it seems that the perspective is not damaged.
-->
I do not know why.
-->
Ready then.
-->
We start from this point here.
-->
First what we are going to do
-->
there is a file called PROM
-->
which is the one that connects to the theme.
-->
Please, the connection line
-->
remove it because
-->
we are going to format that PROM.
-->
Notice that now when you disconnect it
-->
it no longer belongs
-->
to that JOK.
-->
So if there is any problem
-->
to be able to connect it, lower the
-->
JOK a little so that this PROM is free up here.
-->
Later when you reconnect it
-->
there is no problem to be able to
-->
put it in this JOK.
-->
Just in case.
-->
Now we are going to place a file.
-->
This file that I generated here
-->
which is the one I want to integrate
-->
this file that is out
-->
is the one I want to read to integrate it
-->
and send it over there.
-->
This is not finished.
-->
So do the following.
-->
I am going to do
-->
I am going to export this man.
-->
He is an example.
-->
I am going to export this JOK.
-->
And I share this JOK with him.
-->
Perfect.
-->
I am going to share it
-->
here in 10.
-->
Perfect.
-->
And now let me share
-->
this 10 here.
-->
Here is the local project.
-->
Wait a moment.
-->
It didn't generate it correctly.
-->
Wait a moment.
-->
Export item.
-->
Here it is.
-->
JOK connection API.
-->
Perfect.
-->
Perfect.
-->
I am going to compress it.
-->
And I am going to upload it
-->
to the shared folder.
-->
If you wish, go to the shared folder.
-->
I think I have an open Gmail connection here.
-->
I don't know.
-->
I am going to try to send it here
-->
to my team and
-->
it is faster for me to upload it here.
-->
I am going to try to see it.
-->
Done.
-->
Let me download it here in my team
-->
and upload it immediately
-->
to the shared folder.
-->
I already told you how to export it.
-->
To import it.
-->
If it stays there.
-->
Done.
-->
I am going to present your screen
-->
to see how we can import it.
-->
Who called me? Brenda, right?
-->
Yes, Brenda.
-->
I am on your screen.
-->
Go to the shared folder.
-->
Access the Edge.
-->
Open your Edge browser.
-->
I think it should be
-->
in the history.
-->
It should be there.
-->
There it is.
-->
Perfect.
-->
Close that window, please.
-->
From the right side.
-->
That one is perfect.
-->
Now download that one called
-->
Job Connection Open.
-->
Done.
-->
Go to Downloads and unzip it, please.
-->
Now go to the Big Data tab
-->
and it is located in the
-->
Big Data tab.
-->
Right click there.
-->
Give it
-->
Import Items.
-->
Select it, please.
-->
Now I got confused.
-->
Did you have the Open Eye one?
-->
Or was it missing the three?
-->
Which one was missing?
-->
Ok.
-->
Choose the Open Eye tab.
-->
Job Open Eye.
-->
Select the tab.
-->
Wait a moment.
-->
It was missing the other one.
-->
Wait a moment.
-->
Yes. Wait a moment.
-->
Ok. Wait a moment.
-->
I am going to export this one.
-->
This one.
-->
Go to the shared folder again.
-->
Wait a moment.
-->
It is going to be here.
-->
Download the one called
-->
F5, please.
-->
The one called Job Listar 10.
-->
There it is. Job Listar 10.
-->
Less sold items.
-->
And unzip it again.
-->
Perfect. Now go to the Big Data tab again.
-->
Right click there, please.
-->
But wait a moment.
-->
You have to delete it.
-->
Because I think it is called exactly the same.
-->
So delete the one you have.
-->
If you have to close it before deleting it,
-->
you have to close the job.
-->
Close it.
-->
Say no.
-->
And delete it.
-->
There is a small problem with the tool.
-->
Even if I have deleted it,
-->
it won't let me create it with the same name.
-->
Delete it there.
-->
And at the end, it says to delete it.
-->
Delete it from the folder because
-->
it is still in the folder.
-->
Delete it from there.
-->
Exactly.
-->
Delete forever.
-->
Now try again.
-->
From the folder, click in.
-->
Import item.
-->
Look for the download folder.
-->
It should be called Job.
-->
That would be that folder.
-->
Select folder.
-->
Exactly.
-->
Select everything there.
-->
Click on local project.
-->
Select everything.
-->
It is in the AXA folder.
-->
Open it.
-->
Wait a moment.
-->
Perfect.
-->
If you want, go to the AXA folder.
-->
Go to the AXA folder.
-->
It is there.
-->
Go to the AXA folder.
-->
Go to the Aprovesamiento Big Data folder.
-->
And take it.
-->
And upload it to Aprovesamiento Big Data.
-->
And upload it.
-->
And upload it exactly so it stays in the route.
-->
Now delete the folders that
-->
were left there.
-->
Delete it.
-->
And now go to Job.
-->
Click on 10 less sold items.
-->
Finally, copy the whole Job connection
-->
API Open Eye.
-->
It has it there.
-->
Right click.
-->
We are going to copy everything.
-->
The best thing is to right click on the blue part.
-->
Within the blue part of the right click there.
-->
Copy it.
-->
Ready.
-->
And paste it here.
-->
Ready.
-->
It stayed more to the left.
-->
There it is.
-->
Continue more to the left because I need you to delete something.
-->
Select the prompt component.
-->
That prompt that is there.
-->
Select it.
-->
Click there.
-->
And it is going to delete it.
-->
It is going to delete the prompt connection line
-->
to the subject.
-->
Now go down a little bit to the sub job
-->
because later you won't be able to select the prompt.
-->
Go down the complete sub job.
-->
Click on the blue and go down.
-->
No, directly in the blue.
-->
No, in the blue.
-->
There it is.
-->
Well, there it is.
-->
It minimized it. Now it maximizes it.
-->
There in the plus it maximizes it.
-->
In the plus it has the theme 3.
-->
In the theme 3 it was minimized.
-->
In the theme 3, give it more there.
-->
Better go down a little bit.
-->
Select it with the click.
-->
Go down.
-->
And it is going to delete the prompt.
-->
There it is.
-->
Well, let's go then.
-->
If we continue with this part.
-->
This is the window.
-->
Perfect, done.
-->
I told you then.
-->
The data that I want to send to Mr. OpenEye
-->
is in this file that is here.
-->
I already told you why I didn't send it to Duque.
-->
It makes sense because it is very small.
-->
I want to read the data
-->
of this file.
-->
I need a tefile
-->
tefile input.
-->
We add a tefile input delimited.
-->
Well, here each one organizes it a little differently.
-->
I need a theme.
-->
I am going to configure this tefile input delimited.
-->
Remember that the inputs and outputs
-->
need a scheme.
-->
Like the scheme that I need,
-->
I have it in the tefile output.
-->
I go to the tefile output
-->
and I am going to export the scheme.
-->
Everything that is input and output
-->
is going to find this edit scheme.
-->
What is going to vary is
-->
the position in which it is going to find it.
-->
So we give here edit scheme
-->
and we do the export.
-->
In this case it would be the
-->
penultimate button.
-->
I am in the tefile
-->
output delimited.
-->
We are going to call it
-->
menos vendidos.xml
-->
Importante el xml
-->
because then it does not read it.
-->
Venas menos vendidos.
-->
Menos vendidos.xml
-->
We give here save.
-->
And what we do is import
-->
that scheme that is already exported.
-->
We import it in the input.
-->
Edit scheme.
-->
Importamos
-->
y buscamos menos vendidos.
-->
Ahà está.
-->
DÃgame.
-->
Ok, aquà exportar.
-->
Ok, perfecto.
-->
Listo.
-->
Bueno, listo.
-->
Y por último, del tefile output delimited
-->
copie la ruta del archivo de salida
-->
y la va a pegar en el input
-->
porque voy a leer exactamente el mismo archivo.
-->
Entonces voy al output,
-->
copio la ruta del archivo de salida
-->
y es la que
-->
la aquÃ
-->
la borro y la pego aquÃ.
-->
En este caso recuerde que
-->
estos componentes
-->
cuando es un stream debe llevar comillas doble.
-->
Solamente hemos visto
-->
dos casos en que
-->
solamente, perdón, un caso
-->
en que no necesita comillas dobles, que es cuando
-->
creer contextos.
-->
Cuando dice la parte
-->
final. Ah, sÃ.
-->
SÃ, la de salida, sÃ, exactamente.
-->
Aunque de verdad que estoy mal que la salida es
-->
igual a la de salida.
-->
Es diferente a la salida.
-->
Exactamente.
-->
Menos vendidos.xml
-->
pero está en descarga.
-->
No está en temporal.
-->
TendrÃa que buscarlo después
-->
que está ahà en descarga.
-->
Y después
-->
lo importa
-->
en el file delimited.
-->
Parece que le falta agregar el component
-->
de file input, exactamente,
-->
de file input delimited.
-->
Entra al esquema del
-->
file input delimited. Listo.
-->
Ahora por favor vamos entonces
-->
a unir el prom
-->
con el TMAG4. En este caso
-->
a mà me marcó como TMAG4.
-->
No sé si de pronto usted le marcó con el mismo
-->
nombre.
-->
Conectamos el prom
-->
y el
-->
file input delimited con el tema
-->
Como el orden puede
-->
influir con el caso, creo que fue de Luis,
-->
no se fue Alejandro que de pronto lo colocó en un orden
-->
diferente. Para saber
-->
cómo lo vamos a organizar vamos a cambiar el nombre
-->
de la flecha que es importante como buena
-->
práctica. La flecha que conecta
-->
el prom con el TMAG4
-->
la voy a llamar
-->
plantilla
-->
on the record prom
-->
PT
-->
Recuerde cómo cambiarla, le da click
-->
y después nuevamente le da click.
-->
Y la que conecta
-->
el file input con el TMAG4
-->
la voy a llamar resultado
-->
big data.
-->
Listo, perfecto. Ya estamos terminando
-->
esta configuración. No
-->
queda mucho.
-->
Vamos ahora
-->
entonces de aquà donde juega muy importante
-->
para lo que tenga en cuenta a nivel práctico
-->
este señor que habÃa colocado como
-->
porcentaje S. En ese
-->
porcentaje S voy a reemplazar el
-->
valor que me devolvió el big data.
-->
Entonces, ¿cómo hago eso?
-->
Me voy nuevamente a mi talent
-->
mi talent
-->
big data
-->
voy al TMAG4
-->
y lo voy a configurar.
-->
Voy a agregar aquÃ
-->
una variable
-->
el nombre, la variable
-->
la voy a llamar prompt
-->
el nombre y va a ser tipo string
-->
muy pendiente cuando cree esas variables
-->
porque el tipo tiene que
-->
coincidir con la expresión.
-->
Si la expresión devuelve un entero
-->
pues el tipo tiene que ser entero. En este caso la
-->
expresión va a volver un string.
-->
Por último entonces
-->
vamos aquà a lo siguiente.
-->
Hay un pequeño
-->
un momento que no me va
-->
a salir de la forma en que aquÃ
-->
me toca hacer un pequeño cambio. Un momento
-->
a ver si lo hago aquà directamente.
-->
Pero un momento a ver, me toca hacer aquÃ
-->
un pequeño cambio porque
-->
de si lo dejo de esta forma
-->
solamente me va a leer fila por fila
-->
y como hay tres registros me va a
-->
hacer tres llamados a lo que hay.
-->
Me va a tocar hacer aquà un pequeño cambio
-->
para adaptar que este
-->
file input delimited
-->
en realidad
-->
en realidad que este file input delimited
-->
solamente me lea
-->
todos los registros para que envÃe
-->
todos de un solo
-->
llamado al OpenAI.
-->
Me va a tocar hacer aquà una pequeña modificación
-->
disculpa
-->
vamos aquà a la metadata mejor y lo hacemos
-->
acá.
-->
Vaya aquà a file delimited
-->
y va a crear
-->
un archivo delimitado acá en el metadata.
-->
Ahora
-->
salte un momentico Alejandro, ahà vamos a crear
-->
en la metadata que aquà fue un error mÃo aquÃ
-->
inicialmente de
-->
vamos a llamarlo ahora
-->
a este nuevo archivo, vamos a llamarlo
-->
salida big data
-->
big data
-->
vamos a darle aquà next
-->
busque por favor el fichero
-->
donde quedó
-->
creo que en el caso mÃo quedó
-->
sin temporal
-->
se llama producto menos vendido
-->
dale por favor next
-->
no le digas que tiene encabezado
-->
ya le digo, déjelo asÃ
-->
sà dÃgame
-->
ya voy un momento claro que sÃ
-->
listo
-->
antes
-->
se lo voy a compartir inmediatamente
-->
no le coloque encabezado y le voy a decir cuál es el truco
-->
que quiero hacer. Esto lo pude haber hecho
-->
de una forma más complicada
-->
leyendo el archivo pero creo que esta es la forma
-->
más fácil de hacerlo
-->
quiero leer todo el archivo
-->
en un solo llamado
-->
o sea en una sola fila
-->
porque si lo dejo asà por defecto el componente
-->
de este file input me lee fila por fila
-->
ya un momento, ya lo recupero
-->
entonces lo que quiero es que todos los registros
-->
se lean en una sola fila
-->
entonces el truquito
-->
un momentito ya le pongo esto acá
-->
ya lo descargo acá
-->
y lo voy a colocar entonces en talent
-->
td4
-->
listo
-->
puede descargarlo en la carpeta compartida
-->
mire si ya le aparece del f5
-->
¿qué te parece?
-->
déjame para ver
-->
asÃ, ya está, un momento
-->
no es cierto un momento que pasó aquÃ
-->
es el momento que no lo
-->
me la veo y ahora mire para ver
-->
si le aparece o no
-->
ok, por favor, descargue
-->
mire lo siguiente
-->
por defecto dese cuenta que aquà me está leyendo
-->
en dos columnas y en cuatro filas
-->
yo quiero que todo el
-->
contenido me lo lea en una sola fila
-->
¿por qué? porque quiero mandar
-->
en una sola request, quiero mandar
-->
todos esos datos al OpenI
-->
entonces el truco que voy a hacer es lo siguiente
-->
en el separador de campo que está aquÃ
-->
le voy a colocar aquÃ
-->
custom utf8
-->
y le voy a colocar arroba
-->
y en el separador de fila
-->
le voy a colocar custom string
-->
y también le coloco arroba
-->
al cambiar esto
-->
me debe leer todo en una sola fila
-->
lele refresh
-->
y mire como lo lee
-->
lo lee como si fuese una sola fila
-->
este es el truco para poder enviarlo entonces
-->
dele aquà por favor
-->
next
-->
y cambie el nombre
-->
de la columna por resultado
-->
aqui le dejo nuevamente
-->
lo que cambie fue el separador de campo
-->
en custom utf8
-->
y le mande una arroba
-->
y cambie el separador de fila
-->
por una arroba
-->
y tome, entonces que hace eso?
-->
que todos los datos del archivo lo mande en una sola linea
-->
next
-->
aqui
-->
cambielo al resultado
-->
y aqui
-->
finish
-->
borra ahora
-->
al final el tefallinput
-->
que creamos
-->
y tome el tefallinput del metadata
-->
que lo llame
-->
salida big data
-->
y ahora
-->
lo une con el tema
-->
ahora si
-->
ese lo pude haber hecho
-->
en realidad eso que hice lo pude haber hecho
-->
directamente pero lo que pasa es que si lo hacia
-->
directamente sobre el archivo no voy a entender por que lo hizo
-->
al final el movimiento
-->
que hice o la configuracion que hice
-->
fue simplemente para que todos los datos
-->
que me arrojara big data
-->
estuviese en una sola fila
-->
de tal forma que le mandara un solo request
-->
todos los datos al señor
-->
open eye
-->
si digame
-->
menos vendido
-->
el menos vendido
-->
creo que usted no lo tiene
-->
porque usted no lo ha ejecutado
-->
asi como el caso de, no tenia ese archivo
-->
entonces descargue lo por favor, haga lo siguiente
-->
descargue lo del
-->
de la carpeta compartida
-->
que verdad que usted no lo ha ejecutado
-->
todavia, usted no lo tiene
-->
dale este 5 por favor
-->
ahi esta producto menos vendido, ese que esta ahi
-->
descargue lo por favor, si esta en aval de
-->
lo descargue en el temporal
-->
bueno, deje de ir en descarga, no hay problema
-->
vayas entonces al
-->
tal en big data
-->
y busque en descargas ahora si
-->
voy a mostrar como va abriendo de pronto alguno que va por aqui
-->
abra por favor ahi, dale next
-->
ahora cambie el separador
-->
separador de campo
-->
esa codificacion esta bien porque esta en ingles no hay mucho problema
-->
separador de cambio
-->
dele click y escoja la ultima opcion
-->
que es custom utf8
-->
y cambielo para arroba
-->
porque lo cambio por arroba porque el carater
-->
actualmente es punto y coma entonces
-->
para que me de una sola columna
-->
y el separador de fila
-->
dele custom nuevamente custom
-->
y pongale ahi
-->
y cambielo arroba
-->
ese slash n lo cambia arroba
-->
dele un refresh y deberia entonces
-->
traer todos los datos en una sola fila
-->
perfecto
-->
dele next ahora y llame la columna resultado
-->
la columna 0 que esta y llamela
-->
para que tenga mas sentido resultado
-->
dele enter ahora listo dele finish ahora
-->
ahora en el job borre
-->
ese te file input del limited borrelo por favor
-->
ese te file input del limited del job
-->
lo borra si el te file
-->
input del limited que esta y lo borra
-->
suprimir dele con el
-->
tecle suprimir y se arroba listo
-->
ahora vais a la metadata
-->
del file del limited
-->
y busque el archivo que se llama salida big data
-->
lo suelte en el canvas y une
-->
ese señor con el te map
-->
este file input
-->
ahora une el señor
-->
con el te map, cambie el row 7
-->
por resultado de score big data
-->
el row 7
-->
sabe como cambiarlo la linea
-->
nuevamente click
-->
y lo llama resultado
-->
de score big data
-->
big data
-->
big data
-->
enter y ahora enter al te map
-->
y es por donde estamos entonces
-->
listo un momento por favor
-->
le hago aqui el cambio
-->
resultado
-->
under score big data
-->
vamos aqui al te
-->
al te map 4
-->
y ahora si
-->
esta como yo quiero
-->
entonces vamos aqui a la expresion
-->
y recuerde que
-->
para poder modificar ese
-->
porcentaje ese voy a hacer la siguiente funcion
-->
string.format
-->
le va a mandar entonces
-->
la plantilla como primera opcion
-->
y despues le va a mandar resultado big data
-->
entonces como te va a quedar esto en grande
-->
te va a quedar de esta manera
-->
te lo voy a colocar
-->
te va a quedar esto asi
-->
lo que debes copiar es
-->
string.format
-->
ya el nombre de la variable
-->
lo toma directamente
-->
del te block que esta al ladito
-->
en una variable
-->
quien me hablo para guiarla mejor
-->
quien me hablo
-->
Beatriz vamos entonces
-->
exactamente
-->
entonces vamos a ver Beatriz
-->
entra al te map exactamente
-->
dentro de la expresion por favor
-->
dele click en la expresion en la primera
-->
y copia string
-->
con ese mayuscula
-->
string.format
-->
ahi esta el primerito
-->
ahora
-->
reemplaza ese null por plantilla
-->
pero donde el lado derecho estan las variables
-->
mejor
-->
tome la rayada que sale mas facil
-->
doble click ahi
-->
selecciona el siguiente null
-->
doble click en el siguiente null
-->
y vaya el resultado ese que esta ahi
-->
perfecto
-->
ya quedo esta ahi
-->
ok listo
-->
por favor si esta mal yo tambien lo voy a dar aqui
-->
ok
-->
ok
-->
voy a crear aqui una salida
-->
una salida
-->
que la voy a llamar
-->
prompt pt
-->
salida prompt pt
-->
aqui le da en el mas de aca arriba
-->
y manda la variable hacia esa salida
-->
perfecto prompt
-->
y manda la variable
-->
le damos aqui ok
-->
y unimos el tema 4
-->
con el tema 3
-->
es decir entonces vamos a unir este señor
-->
por la salida
-->
dele click derecho porque despues no le sale
-->
situa en el tema click derecho
-->
y debe aparecerle cuando aparece en fila
-->
debe aparecerle la salida que configuro
-->
tome esa salida prompt
-->
y la suelta en el tema 3
-->
listo ahi
-->
tema 3 la suelta
-->
ya nos falta ya una confersion
-->
sencilla que es la del tema 3
-->
voy aqui al tema 3
-->
y en realidad
-->
aqui
-->
listo modifico lo siguiente
-->
aqui como no tengo los nombres de
-->
prompt
-->
payload
-->
rose
-->
la madre rose
-->
expande la expresion esta por favor
-->
mire por favor como llamo
-->
este es lo importante de colocarle los nombres
-->
a estas filas
-->
mire como el nombre de esta fila que esta aqui
-->
la segunda que esta aqui en el caso mio
-->
rose
-->
entonces nada mas cambio
-->
aca tengo rose 5 y cambio rose 5
-->
cambio el 5 por 6 para que lo una
-->
mire como se llama aca
-->
y cambia el encabezado
-->
porque la variedad se llama igual
-->
rose 6
-->
y cambia
-->
momento momento momento
-->
perdon no no es rose 6
-->
es la de arriba
-->
es que
-->
haga lo siguiente
-->
para no equivocarlo
-->
con estos nombres me equivoque
-->
vamos a nombrarlo bien
-->
cierre un momento y vamos a nombrarlo bien
-->
el payload
-->
que une con tema 3
-->
cambie ese nombre a payload
-->
para ubicarme bien
-->
la linea que une
-->
payload con tema 3 cambiela a payload
-->
si si
-->
payload
-->
para ubicarme
-->
mejor
-->
que ese nombre road 1 road 3
-->
te pega una vez en una partida 1
-->
road payload listo
-->
y ahora si vamos a entrar a configurarlo
-->
bien aqui esta bien
-->
y en realidad
-->
la segunda
-->
la segunda columna no es payload
-->
sino prompt
-->
que seria
-->
la de arriba
-->
prompt
-->
ahora si prompt
-->
porque el payload va aca y el prompt va de segundo
-->
para verificar
-->
que todo esta bien
-->
deberia aparecer 2 lineas
-->
uniéndose a la variable
-->
aqui esta una linea amarilla
-->
de la parte abajo y una linea amarilla de la parte arriba
-->
si no estan esta mal configurado
-->
solamente cambie
-->
la segunda
-->
encabezado
-->
lo cambia a promptpt
-->
y yo creo que ya
-->
verifique por favor cuando termine aqui
-->
que las 2 filas
-->
las 2 entradas
-->
se unan en la variable con una linea amarilla
-->
diciendo que estan relacionadas a esas variables
-->
la entrada con la variable
-->
vamos a darle aqui ok
-->
listo
-->
no lo hago usted aca lo voy a hacer solamente yo aca
-->
voy a agregar aqui un telo road
-->
para que me muestre como quedo el payload
-->
anterior
-->
y ya lo ultimo que va a hacer
-->
usted nada mas alla es unir
-->
los subjogs
-->
para que se ejecute primero el de arriba y despues el de abajo
-->
como unimos los subjogs
-->
va al primer elemento del
-->
subjog 1 que quiere unir
-->
y lo une con el
-->
disparador o un subjog ok con el primer
-->
componente del subjog 2
-->
en este caso primero tiene que ejecutar
-->
el input de arriba
-->
para que de la salida y despues
-->
se ejecuta hacia lo que hay
-->
voy para ver
-->
vamos a ver
-->
ya se rojo
-->
perfecto
-->
la pregunta es que hay veces que
-->
se queda
-->
por la conexion
-->
bueno voy a probar
-->
voy a ejecutar
-->
voy a comprobar que este sea el jog
-->
listar 10 articulos
-->
vamos a ver
-->
vamos a ejecutarlo inicialmente para ver
-->
y digo para ver
-->
si hay algun error aqui
-->
bueno
-->
empezó haciendo la parte de big data
-->
leyendo directamente
-->
del dejado vamos a ver
-->
empezó a hacer
-->
el pron listo
-->
ah
-->
que pena
-->
un momento
-->
se nos olvido cambiar
-->
el archivo
-->
pron que lo modificamos creamos uno nuevo
-->
se nos olvido
-->
se nos olvido a mi
-->
vayase al pron
-->
pron pt el archivo
-->
y note que esta
-->
direccionando al pron del dÃa de ayer
-->
hoy hicimos un nuevo archivo
-->
para cambiarlo nada mas vaya aqui donde dice
-->
los 3 punticos al final
-->
y cuando le aparezca este
-->
dele chain
-->
no le de un day sino chain
-->
y el le va a permitir cambiar
-->
la ubicacion del archivo
-->
nuevamente aqui le doy
-->
despues que doy chain
-->
tengo que dar nuevamente click
-->
y busco el nuevo pron
-->
que yo lo llame
-->
pron articulos menos vendidos
-->
listo
-->
porque estaba
-->
ejecutando el pron anterior
-->
vamos a ver ahora con este nuevo pron
-->
a ver si
-->
si aqui lo mando
-->
no chain
-->
quien me dijo
-->
Brenda fue Brenda
-->
voy a para que lo muestre
-->
dele chain primero
-->
y listo
-->
ahora escoja el nuevo pron
-->
que seguramente lo tienen
-->
credenciales open eye
-->
ese menos vendido listo
-->
dele abrir
-->
y dele ejecutar usted inicialmente
-->
para ver si no tiene ningun problema
-->
tengo un pequeño error
-->
el pron esta bien
-->
pero tengo un error
-->
vamos a ver si a todo le sale el mismo error
-->
me quede con la ventana aqui de Beatriz
-->
para ver si le sale el mismo error
-->
hay algunos que ya estan ejecutando
-->
vamos a ver
-->
vamos con Brenda tiene un error
-->
exactamente
-->
dele ok
-->
para buscarlo
-->
vaya donde dice code
-->
vaya al code del job
-->
y debe marcarle algun punto rojo
-->
que esta mal configurado
-->
payload dice que no existe eso
-->
entonces vayamos
-->
eso esta en el tema
-->
vayas al designer
-->
y entre al tema que une
-->
ese tema 4 que esta ahi
-->
ah no perdon es en el 3
-->
bien
-->
dese cuenta lo siguiente
-->
lo que pasa es que
-->
en primer lugar dese cuenta que row 8
-->
no esta en amarillo verdad
-->
y no esta
-->
no esta porque
-->
porque ahi su conexion se llama
-->
row 8
-->
significa en pocas palabras lo siguiente
-->
vamos a hacer lo siguiente
-->
dele ok
-->
dele ok mejor ahi
-->
y vamos a cambiar el nombre de la linea
-->
dele ok
-->
y la linea que conecta
-->
baja un poquito por favor
-->
esa linea que conecta ese row 8
-->
que esta ahi cambia el nombre a payload
-->
por favor si estan amable
-->
ahora si entre al tema 3 y asi nos vamos a ubicar
-->
listo tema 3
-->
doble click por favor listo
-->
entre si estan amable
-->
ahora si
-->
dese cuenta que ahora si le marco bien
-->
payload y prom ahora si esta bien
-->
dele ok vuelve a ejecutar
-->
parece que ese nomas le da error
-->
vuelve a ejecutar nuevamente
-->
vamos a ver que error le sale a usted
-->
empezó ya
-->
la parte de big data
-->
por donde va
-->
se esta demorando
-->
por donde va
-->
subo un poquito por donde va
-->
esta por ahi
-->
que raro que haya demorado en ese punto
-->
note que ahi
-->
aparece starting
-->
si nota que esta en ese punto
-->
ahi que se ha demorado que raro
-->
que se haya demorado de esa forma
-->
espera un momento para ver si de pronto
-->
ahora vuelvo con usted
-->
no se si alguno le salio
-->
y que le salio
-->
muestre para ver
-->
ya voy para ahi
-->
listo vayase al designer
-->
porque le si copio el
-->
subo por favor
-->
vayase al tema 2
-->
tema 2
-->
ese que esta ahi
-->
no se porque esta ahi
-->
pero bueno
-->
note el product line de
-->
hay dos tablas
-->
la izquierda y la derecha que esta en la parte de abajo
-->
la tabla de la derecha
-->
le dice los archivos de salida
-->
ese product line tiene que ser a intero
-->
no se porque se cambio a este
-->
cambielo ahi a intero
-->
que es tan amable
-->
dele enter
-->
listo dele ok
-->
y ahora vuelve a ejecutar para ver
-->
Brenda detenga ese job
-->
mate los kills y vuelve a ejecutar nuevamente porque esta como rara
-->
ya no esta que demora tanto
-->
bueno creo que hay algunos listo vamos a ver
-->
ok dele ok un momento ahi
-->
ok
-->
entra nuevamente al tema 2 a ver que paso ahi
-->
porque hay una conversion extraña al tema 2
-->
no que pena ahi me equivoque
-->
fue yo ahi en realidad el stream estaba bien
-->
deje el stream por favor
-->
deje yo acceder un momento
-->
que fue rovino
-->
un momento enter
-->
ok ahi nuevamente
-->
yes no el cuento
-->
era con
-->
ya se
-->
accedo un momento
-->
espere un momento que accedo al studio
-->
porque el problema
-->
creo que con el tipo de datos no era doble
-->
era que era
-->
un momento a ver
-->
este señor que esta aqui
-->
es
-->
total doble no veo ningun problema
-->
permÃteme verlo
-->
el error está
-->
en
-->
dice que es stream ah ya
-->
aqui hay algo
-->
extrañamente raro ya se que le pasa
-->
este
-->
esta fila tenia que mandarla
-->
aca
-->
para cambiar la posición va a colocarla de primero
-->
simplemente por cuestión de gusto
-->
ok
-->
y no se porque
-->
en ese job pasa esto aqui
-->
no era pro online sino
-->
era rovno voy a guardar
-->
y verificar que el cambio
-->
esta aqui listo no hay ningun error
-->
voy a proceder a ejecutar
-->
perfecto listo este es el error que tengo
-->
un cuarto ciento de menos deja ahi por favor
-->
brenda usted tiene la máquina virtual arriba
-->
que pronto toda tiene arriba
-->
ah un momento ya voy aqui
-->
prendo el caso suyo
-->
se quedo ahi verdad
-->
permita un momento voy a acceder a su máquina de manera
-->
interactiva y voy a comprobar un par de
-->
listo voy a ver algo
-->
aqui no es que la máquina no tiene encendida
-->
entonces intentaba conectarte a esa máquina
-->
y por eso se bloqueo
-->
vamos a encenderla aqui iniciar y
-->
voy a bloquear un momento
-->
listo ya esta esperamos un momento
-->
vamos a ver vamos a ejecutar entonces
-->
ahora si nuevamente por haber
-->
ahora si vean este cuenta que va por aqui
-->
listo este era el error
-->
que queria que no saliera listo
-->
quien mas tiene errores
-->
diferentes a este cuatrocientos que nos devolvio
-->
nadie mas listo voy a corregir este error
-->
cuatrocientos que todavia no se que vamos a ver
-->
cuatrocientos que dice
-->
ahhhh
-->
este es el de Beatriz este es el mio
-->
a ver el
-->
cuatrocientos por que se da
-->
en pocas palabras me esta diciendo que mande
-->
de forma RadalJason
-->
este es el payload que estoy enviando
-->
ah listo lo que pasa es que este señor
-->
OpenAI es un poquito sensible
-->
que tiene que tener en cuenta lo siguiente
-->
dese cuenta que yo modifique
-->
la lectura del archivo
-->
pero no modifique el archivo
-->
como tal el archivo es de salida big data
-->
y parece ser que
-->
este señor no le gusta
-->
los saltos de linea
-->
entonces
-->
que vamos a hacer
-->
vallase al tema
-->
parece que es eso vallase al tema 3
-->
perdon al tema 4
-->
perdon al tema 4
-->
perdon al tema 4
-->
vallase a la expresion del tema 4
-->
vallase a la expresion del tema 4
-->
dele aqui al final
-->
un punto replace
-->
punto replace
-->
ya le coloco
-->
como queda en grande
-->
va a agregar esto
-->
al final
-->
un punto replace
-->
abro parenthesis, comilla doble
-->
slash n y reemplazo los
-->
slash n por
-->
espacio en blanco
-->
que al señor no le gusta
-->
seguramente en el jason que vaya un
-->
salto de linea
-->
porque bueno ya es cuestión muy propia
-->
de el de porque no debe aceptar eso
-->
agrega al final un punto
-->
replace
-->
mientras usted
-->
voy ejecutando para ver si ese es el error
-->
creeria que ese es el error
-->
perfecto
-->
me respondio debe de haber la respuesta
-->
en el correo voy a confirmar
-->
para ver
-->
efectivamente ya me respondio
-->
para aumentar la mente de productos entre hombres en yangon
-->
es importante tener en cuenta la parte cultural
-->
investigar las preferencias y tendencias de consumo de hombres en yangon
-->
bueno ahi el pron si estuvo mal
-->
la idea es saber como integrar
-->
bueno a quien no le funciono a ver
-->
me regala dos minutico y busco un cafecito
-->
y mientras usted va
-->
ejecutando ahi busco un cafe ya vengo
-->
bueno volvi que pena
-->
cuenteme si a quien no le funciono
-->
si le funciono a todos
-->
ah ok
-->
si creo que
-->
no se si le canso
-->
vaya a ser al tema 4
-->
debe tenerlo como tema 4
-->
este que esta ahi
-->
tema 4
-->
por cierto
-->
para cambiarle el nombre tambien a los componentes
-->
que de pronto
-->
con tema 4
-->
hace lo mismo como cambiamos las flechitas
-->
le da click en el componente
-->
que quiere cambiarle el nombre en el label
-->
click nuevamente
-->
y le puede cambiar el nombre
-->
si acaso
-->
vallense a la expresion por favor si es tan amable Brenda
-->
ahi falto
-->
listo
-->
al final del parentesi coloque un punto
-->
replace ahi esta el primero del enter
-->
abra comilla doble ahi
-->
slash n al
-->
el otro
-->
el otro el otro el
-->
ese
-->
cierre comilla doble
-->
y ahora esta el 0
-->
coloque comilla doble
-->
espacio comilla doble para reemplazar ese caracter ahi
-->
espacio comilla doble
-->
dele ok
-->
por favor si es tan amable
-->
dele ok
-->
cierre la entana y vuelve a ejecutar para ver
-->
listo ya esta por la linea
-->
consumiendo listo
-->
rese su correo por favor
-->
bueno ya claro que el prod
-->
no fue el mejor pero la idea es
-->
creo que ahi tenemos el ejemplo
-->
de como podemos combinar todo
-->
resultado con la parte de ahi
-->
lo que tendria que mejorar es el prom
-->
y pues evidentemente saber
-->
como manejar los datos tiene que tener en cuenta que cuando
-->
manda un json ese open ahi es muy
-->
celoso con algunos caracteres
-->
ya iremos con el caso del
-->
salto de linea y puede haber unos caracteres
-->
que de pronto no los acepta
-->
si le llego
-->
Brenda ok
-->
listo
-->
por un lado y nos vamos con Hai
-->
explico entonces
-->
no vamos a lo siguiente mejor
-->
mientras voy explicando
-->
porque que no se si tenemos memoria
-->
voy a cerrar este señor por favor
-->
abra
-->
algo que se llama
-->
el docker de stock
-->
se busca aqui docker aqui esta no aqui esta
-->
esta en el escritorio dice docker de stock
-->
le da doble clic por favor
-->
para que abra la interfaz
-->
listo lo deja ahi que vaya abriendo un momento
-->
la interfaz mientras yo voy explicando
-->
como la arquitectura que estamos manejando
-->
actualmente
-->
listo mientras eso va subiendo mire mi pantalla
-->
un momento y le voy a hacer el diagrama de lo que tenemos
-->
entonces
-->
diagrama aqui
-->
en blanco perfecto
-->
bien veamos por un lado
-->
tenemos la maquina
-->
virtual que es la de windows
-->
listo esta maquina virtual
-->
tiene direccion ip
-->
estatica perdon
-->
ip
-->
10.0.3.15
-->
sobre windows
-->
tenemos
-->
otra maquina virtual con hyper v
-->
que se llama
-->
hadut y que esta maquina virtual
-->
tiene direccion
-->
10.0.3.16
-->
ahora sobre
-->
windows tengo
-->
instalado el docker que es
-->
un motor de contenedores
-->
y sobre docker
-->
tenemos montado lo siguiente
-->
tengo montado tres contenedores
-->
un contenedor de
-->
postgres que ya les digo para que sirve eso
-->
postgres sql
-->
un contenedor que se llama
-->
metastore
-->
metastore
-->
y un componente
-->
que es el que tiene
-->
instalado el hi explico entonces
-->
inicialmente
-->
este contenedor no tiene direccion ip
-->
aunque se le puede configurar pero normalmente
-->
la ip seria
-->
la misma ip de la maquina que este el señor
-->
instalado el contenedor
-->
explico entonces
-->
como es la historia ahora
-->
en primer lugar hi
-->
es un motor ahora
-->
pensado para bodegas de datos
-->
que utiliza
-->
tecnologÃa de big data
-->
hi va a ser el almacenamiento
-->
en un cluster de hadu
-->
entonces ya esto esta configurado para que este
-->
hi guarde los datos
-->
en la maquina virtual en pocas palabras
-->
cuando creemos tablas insertemos
-->
registros ahi van a quedar guardados
-->
en el cluster de hadu que esta a este
-->
lado
-->
para que es ese postgres
-->
resulta en que lo que
-->
voy a hacer es que voy a
-->
colocar la misma capa
-->
sql que usted conocio
-->
pero al final cuando yo hago una consulta
-->
con un sql insert
-->
al final lo que estoy usando es ya sea
-->
macredius o ya sea sparr
-->
para hacer cualquiera de las cosas
-->
entonces para poder simular
-->
el sql insert
-->
el create table
-->
debo utilizar un motor de base de datos
-->
en este caso el motor que usamos
-->
el que estoy usando es postgres
-->
por lo tanto este señor va
-->
a tomar la sintaxis de pokres
-->
para crear tablas
-->
y como las creas a través
-->
de un elemento intermedio que se llama
-->
metastore en pocas palabras
-->
la estructura de datos
-->
de tu base de datos va a
-->
quedar almacenada aqui en postgres
-->
pero los datos
-->
van a quedar almacenados no en esta instancia
-->
de pokres sino en este hadu de aqui
-->
cuando hago una consulta
-->
en select aterisco
-->
por la tabla que tu quieras lo que va a ser
-->
entonces hay es que va a convertir eso
-->
o ya sea a macredius
-->
o ya sea sparr
-->
en el caso de versiones 3 hacia abajo
-->
haria y convertiria
-->
osea practicamente lo que hace es
-->
big data
-->
es la unica diferencia que el select aterisco
-->
que es una postgres sql me la va a convertir
-->
en un macredius
-->
y si estoy de 3 hacia arriba
-->
puede que la consulta
-->
me la haga con spark
-->
porque digo que puede
-->
porque tu puedes configurar el high
-->
en version 4
-->
para que ejecute ya sea con macredius
-->
o con spark como tu desees
-->
entonces partiendo ese hecho
-->
esta seria la arquitectura
-->
vamos a comprobar
-->
lo siguiente entonces vamos a
-->
la maquina virtual a ver si ya me subio
-->
el docker de esto
-->
ok aqui esta
-->
por favor
-->
por defecto estan los
-->
3 contenedores
-->
seleccionados
-->
si digo iniciar puede
-->
que me daña el contenedor porque tienen que haber
-->
un orden en los cuales vamos a subir los contenedores
-->
lo vamos a subir en el siguiente orden
-->
primero
-->
seleccionado todos los contenedores
-->
y el primero que vamos a subir
-->
se llama high postgres
-->
que es el que tiene instalado el postgres
-->
ahi van a quedar las tablas
-->
dele por favor seleccionale
-->
y da click aca en el play para iniciar
-->
listo el verde indica
-->
en que subio de manera
-->
correcta ahora suba por favor
-->
el metastore
-->
que seria el segundo
-->
lo selecciona y le da play
-->
para iniciar
-->
y no hay ningun problema si digame
-->
ok creo que hizo esto
-->
creo que esta asi mas o menos
-->
esta aqui
-->
aqui le queda como una pequeña flecha
-->
que esta aqui no se si me parece
-->
si no le da click
-->
ah listo perfecto
-->
verifique que subo
-->
ahora el segundo por favor
-->
por ultimo subamos ahora el de high
-->
que seria el ultimo el primerito que aparece
-->
que pasa
-->
si aqui
-->
me salio un pequeño error
-->
no se si a usted le subio
-->
no subio
-->
asi no subio
-->
hagamos lo siguiente entonces
-->
vamos a volverlo a recrear porque seguramente paso algo
-->
en alguna configuracion no se
-->
de la maquina virtual
-->
de windows
-->
hagamos lo siguiente estando los 3
-->
seleccionados primero
-->
detengalos aqui
-->
en el cuadrado ese
-->
detener
-->
y cuando termine de
-->
tenerlos por favor
-->
le va ahora
-->
en delete
-->
para borrar los contenidores
-->
delete forever
-->
aclaro lo que pasa es que
-->
si me demo
-->
me podia demorar mas
-->
corrigiendo el error porque que pasa
-->
que volviendolo a crear
-->
entonces vallase al proveedor de archivo si esta amable
-->
y en el proveedor de archivo
-->
en la carpeta de descargas
-->
va a encontrar una carpeta
-->
que se llama docker high
-->
2.0
-->
si la vio dele doble click por favor
-->
permiteme confirmar algo aqui
-->
si perfecto esta bien
-->
docker high 3.0
-->
ahora en esta carpeta
-->
que esta aqui
-->
vamos a decir este señor me deja
-->
no no me deja en esta version este es windows
-->
entonces hago lo siguiente sencillo
-->
en la barra de direcciones
-->
que esta aca arriba le va a dar click
-->
un poquito
-->
en la misma barra pero
-->
que no seleccione ninguna letra
-->
de tal forma que seleccione todos los elementos
-->
y escriba cmd
-->
y le da enter
-->
y me va a abrir entonces una terminal shell
-->
de cmd
-->
en esa carpeta que he escogido
-->
y aqui vamos a escribir lo siguiente
-->
docker
-->
compose
-->
deja y se lo escribo aca entonces
-->
mejor lo copio
-->
un momento lo copio de aqui
-->
y para que usted lo vea mejor
-->
aca va a copiar
-->
esto exactamente asi docker
-->
compose op-d
-->
le das enter por favor
-->
y te va nuevamente al docker
-->
de estos
-->
y confirma que tenga los tres
-->
habilitados a poner el comando
-->
claro como lo puedo perfecto dice
-->
Fernanda voy a ver para ver que paso
-->
en el comando y te digo que paso que salio
-->
vamos a ver un pequeño
-->
un momento a ver
-->
falto el op
-->
falto op despues de compose
-->
op up
-->
dale flecha arriba dale flecha arriba
-->
flechita arriba para que te vuelva
-->
para que te coloque el comando
-->
y tu le dices
-->
nada mas
-->
entonces devuelve con flecha a la izquierda
-->
espacio y dale enter ahora si
-->
espera un momento
-->
listo vete ahora a tu docker
-->
listo
-->
perfecto
-->
claro que no se que paso inicialmente
-->
ahora a lo siguiente para
-->
no entrarle mucho comando y vamos directamente aqui
-->
vallese por favor aqui
-->
al que dice high server
-->
dele click a high server
-->
y ahora por favor dele click
-->
ahora
-->
aqui hay un menu
-->
vaya donde dice
-->
exec de ejecutar
-->
a mi particularmente
-->
no me gusta esta shell
-->
me gusta mas coloque la palabra
-->
bash esa shell no me gusta
-->
mucho coloque bash
-->
para cambiar de shell
-->
me gusta mas esa
-->
esa shell no me gusta
-->
entonces vamos listo bash
-->
bash
-->
ahora del siguiente comando
-->
del comando
-->
high enter
-->
tiene que aparecerle ahora un pron llamado
-->
beeline
-->
que es el
-->
el driver
-->
que conecta hacia high
-->
aparece beeline a todos
-->
beeline
-->
ahi esta
-->
ahora vamos a conectarnos a la base de datos
-->
como nos conectamos a la base de datos
-->
tiene que ir al siguiente comando
-->
slash connect
-->
un momento aqui
-->
slash connect
-->
y entonces le da
-->
jdbc
-->
dos puntos
-->
high dos cero
-->
dos punto
-->
slash slash localhost
-->
asi que
-->
mejor hagamos lo siguiente
-->
le agrego este comando mejor
-->
en la carpeta compartida para que no tenga que escribirlo
-->
dos punto diez mil
-->
diez mil
-->
aqui espera un momento y lo coloco en la carpeta compartida
-->
este comando
-->
si de admiracion perfecto
-->
si de admiracion
-->
si quiere lo voy a colocar en la carpeta compartida
-->
mejor para que no este
-->
copiando eso
-->
porque voy a hacer esto
-->
porque dentro de talent
-->
talent puede hacer muchas cosas con base de datos
-->
lo unico que no hay un componente directo
-->
para crear una base de datos, entonces voy a crear aqui
-->
una base de datos
-->
espere un momentico y lo subo aqui en talent
-->
en la carpeta expresiones que habiamos
-->
usado ya esta
-->
voy a la carpeta
-->
compartida aca y la voy a copiar mejor de ahi
-->
a ver
-->
seria one drive
-->
aqui tengo tv4
-->
expresiones
-->
copia el comando
-->
y lo pega directamente aca
-->
en la terminal
-->
deberia permitir copiarlo
-->
perfecto ese es el comando
-->
si esta mal
-->
colocalo en edge
-->
ah perdon
-->
no tiene enter
-->
que pena
-->
no tiene
-->
no tiene usuario
-->
al final te va a aparecer un prompt
-->
que es la linea de conexion
-->
hacia el motor
-->
vete por favor ahi
-->
si quiere escribe ahi one drive
-->
que seguramente ya lo has usado
-->
ahi te aparece un momentico
-->
ya voy a tu pantalla
-->
listo estas ahi
-->
entre expresiones
-->
si esta mal es fernando
-->
copia de la ultima linea
-->
la vas a pegar en el shell
-->
exactamente ahi
-->
click derecho
-->
paste ahi
-->
listo
-->
enter
-->
no tiene usuario
-->
enter y tampoco tiene paso
-->
entonces enter
-->
y ahi estamos conectados
-->
listo entonces
-->
ahora veamos un par de comandos
-->
si digame
-->
ah
-->
quien me hablo Beatriz si
-->
y vamos a ver que paso vamos a ver
-->
porque el comando no funciono
-->
vamos a ver
-->
ah primero dele hay enter
-->
y va a aparecer un prompt que dice
-->
beeline espera un momento por favor
-->
listo perfecto ahora
-->
copia la linea osea arriba
-->
tiene ya el comando listo
-->
copielo por favor donde dice
-->
ah donde dice
-->
conect jdbc
-->
dale flecha arriba parece que le coge el comando flecha arriba
-->
para ver no no un momento no no
-->
salio salio salio dele hay nuevamente
-->
sabe porque no le funciono
-->
porque mando mil ira diez mil
-->
esta escuchando por el puerto diez mil
-->
permÃtame yo le intento a ver un momento
-->
yo se lo copio para ver eso
-->
dele hay ahi asi dele enter
-->
y voy a intentar copiar este comando aqui
-->
pero un momento le intento copiar
-->
este comando
-->
seria este de aqui
-->
y lo voy a pegar aca
-->
el problema es que era diez mil
-->
ahora si
-->
por limita de un cero nada mas
-->
listo seguimos entonces
-->
a ver vamos a dar los siguientes
-->
comandos entonces
-->
vamos aqui a hacer
-->
ah yo si estaba buscando
-->
y dije cari si listo si gracias
-->
ah perdido
-->
primer comando dele el siguiente
-->
deben terminar en punto y coma los comandos
-->
select
-->
curren underscore
-->
database deberia devolverle
-->
punto y coma deberia devolverle el nombre
-->
curren aqui que pena de curren
-->
curren asi
-->
actual base de ahi
-->
le confirmo entonces
-->
confirmo aqui
-->
para ver
-->
espere para ver en que me
-->
equivoque en el comando creo que era
-->
curren me parecio que era curren el comando
-->
dejai
-->
confirmo para ver ah listo
-->
el comando necesita terminar
-->
con parentesis
-->
creo que es mejor
-->
estrategia que lo haga en uno pad para poder
-->
copiar y pegar porque no deja aqui
-->
ahi asi
-->
confirmo
-->
para ver
-->
aqui me devolvio
-->
ahora si
-->
perfecto ahi esta
-->
en la base datos default
-->
el comando es con
-->
listo ahora si llevamos de rapide
-->
que estamos un poquito atrasados vamos entonces a hacer lo siguiente
-->
vamos a crear nuestro propio base datos
-->
create database
-->
axa
-->
para conectarse axa le da use
-->
axa recuerda de terminar en punto y coma si no sale
-->
vamos a ver
-->
vamos a ver
-->
vamos a ver
-->
vamos a ver
-->
dale un
-->
select
-->
curren database
-->
ah
-->
select curren
-->
ah porque tienes
-->
data
-->
listo cambio
-->
entonces espera un momento a ver
-->
data
-->
ah
-->
si porque que si
-->
no tuviese conectado
-->
ahi esta bien conectado perfecto
-->
vamos a hacer los siguientes pasos entonces
-->
dame gracias a ti listo voy a salir de aqui
-->
creamos la base datos recuerda de terminar
-->
en punto y coma use axa
-->
vamos a crear aqui
-->
una tablita sencilla
-->
create table
-->
create table
-->
aqui la ventaja del meta store
-->
es que puedes usar la sintaxis de
-->
de postgres
-->
y tambien una sintaxis de ya
-->
por ejemplo voy a colocar aqui empleados
-->
abro parentesi de aqui
-->
voy a crear aqui identificación
-->
de tipo integer
-->
y voy a colocar aqui
-->
nombres
-->
integer aqui que vena
-->
integer
-->
nombres bar chart
-->
de 60
-->
tambien puede ver usados stream
-->
punto y coma
-->
y aqui aplicas
-->
exactamente ya lo que sabes de base
-->
de datos
-->
vamos a insertar aqui un registro
-->
insert quinto
-->
empleados
-->
values
-->
abro aca
-->
y le digo este señor
-->
como es entero necesita comida simple 92
-->
y el nombre suyo
-->
si usted quiere John aqui y termina
-->
con punti com si
-->
o si usted quiere si
-->
al final lo que necesito
-->
nada mas es esta base de datos
-->
si quiere yo hago lo demas para que lo vea
-->
creo que ahi me va a mostrar
-->
nada mas
-->
necesito que cree la base de datos
-->
listo
-->
entonces permiteme yo creo la tablita aca
-->
porque la tabla la vamos a crear ya directamente
-->
aqui lo que necesitaban nada mas era crear
-->
la base de datos
-->
pero permiteme hacer un ejemplo aqui
-->
y voy a insertar entonces
-->
voy a crear la tabla
-->
crear la tabla
-->
copiar
-->
listo copiar
-->
creo la tablita aqui
-->
si señor
-->
si señor
-->
y no tenemos
-->
y en realidad tengo poca memoria
-->
osea con el
-->
docker tengo menos memoria
-->
que la que tenia anteriormente
-->
si señor efectivamente
-->
recordara que normalmente
-->
cuando yo tengo una base de datos
-->
normalmente
-->
los procesos de actualizacion no hay
-->
mas que todo hay procesos de insertion
-->
creo que no me creo
-->
la tabla que fue
-->
voy a darle aqui show tables
-->
para que me muestra las tablas
-->
show tables
-->
punto y coma
-->
si empleados
-->
ah me equivoque yo
-->
para me equivoque yo fue
-->
empleados
-->
si efectivamente
-->
y creo que bueno ya Alejandro lo tiene
-->
nota que ahi te aparece un mac reduce
-->
cierto
-->
significa que para insertar esta utilizando
-->
mac reduce
-->
sobre el nodo de
-->
de clusters de hadut
-->
entonces
-->
si me voy a este nodo ahi esta
-->
ahi esta exactamente
-->
si me voy al nodo este
-->
voy a buscar el nodo
-->
el cluster en el
-->
en el browser
-->
un momento a ver
-->
voy a ver aqui
-->
browser file system
-->
y note que aqui tiene ya un
-->
propietario hype de dos carpetas
-->
opt y ten
-->
indicando entonces
-->
confirmando que en realidad
-->
tiene que subir de ax al root
-->
hay que subir de nivel
-->
y van a notar entonces que ahi creo
-->
unas carpetas el usuario hype
-->
que es donde tiene almacenado
-->
los datos como tal
-->
ah bueno eso es un
-->
interesante el como
-->
depende de la tabla de como
-->
se definen hype
-->
entonces para esas tablas vamos a
-->
subir para adelantarnos
-->
esa pregunta deja la abierta
-->
y te la respondo ahora cuando creemos la
-->
tabla aca con el hype
-->
vamos entonces
-->
aca listo
-->
lo unico que necesitaba era crear al final
-->
la tabla
-->
perdon la base de datos
-->
y confirmamos aqui
-->
que ya me esta insertando en
-->
en el en la maquina virtual
-->
del cluster dejado aqui no
-->
voy a hacer mas nada entonces me voy a meter aca con
-->
el talent big data
-->
voy a cerrar todos estos señores de aqui
-->
bueno perdon deja uno abierto
-->
porque voy a copiar esta conexion de aqui
-->
no mentira en este ejemplo pequeñito
-->
no
-->
en este ejemplo pequeñito para que usted entienda como se hace eso aqui
-->
bueno listo entonces
-->
voy a crear una subcarpeta en axa que se llame
-->
hype
-->
perfecto
-->
y vamos entonces aqui en
-->
hype vamos aqui crear
-->
job por favor crear job
-->
dentro de la carpeta de hype
-->
coloque aqui job
-->
creacion
-->
tabla
-->
hype
-->
listo voy aqui donde dice
-->
finish
-->
y voy a hacer entonces lo siguiente
-->
este ejemplo voy a tratar de
-->
combinar varios porque el
-->
tiempo no nos esta dando vamos entonces
-->
vamos a crear una conexion
-->
que no la hemos hecho a una base
-->
de datos
-->
hype es una base de datos siendo que utiliza
-->
la tecnologia de big data para progresar sus datos
-->
pero es una base de datos
-->
entonces para hacer conexiones hacia base de datos
-->
me voy aqui donde dice debe connection
-->
click derecho
-->
crear conexion
-->
vamos a llamarla entonces
-->
axa
-->
por costumbre es mas costumbre
-->
mÃa por experiencia
-->
los nombres de conexion de base de datos
-->
le doy el nombre de base de datos
-->
y coloco dentro de parentesis el motor que la tiene
-->
esto por experiencia
-->
que tengo desde algunos procesos
-->
que he tenido que a veces me tengo
-->
dos bases de datos con igual nombre pero a veces me pierdo de ahÃ
-->
a que motor me estoy refiriendo
-->
entonces para costumbre por buena
-->
practica mÃa le coloco axa
-->
parentesis hype
-->
le coloco el nombre de base de datos y dentro de parentesis
-->
le coloco el motor en el que
-->
esta
-->
vemos la key next por favor
-->
y alto seguido lo que va a hacer es que vas a
-->
buscar aqui donde dice debe tie
-->
el driver de conexion
-->
hacia la base de datos que necesitas en este caso
-->
expande y me interesa el que
-->
se llama hype
-->
aqui hay un pequeño bug
-->
con esta pequeña
-->
bueno no
-->
login no tiene
-->
servidores localhost
-->
porque aunque esta en el contenedor
-->
recordara que el contenedor tiene
-->
la misma hype de la máquina en que este
-->
entonces queda localhost
-->
el puerto queda igual lo unico diferente
-->
es que voy a conectarme no a base de datos default
-->
sino a base de datos que se llama axa
-->
como tu lo hayas llamado
-->
tienes que bajar el scroll para que te de la conexion
-->
dale un
-->
text connection aqui por favor
-->
ok
-->
este exportar como
-->
el contexto lo que hace es que te crea las variables
-->
y te las porta como contextos recuerda que
-->
los contextos son las variables de entorno
-->
cuat y de produccion
-->
que ya hicimos una el dia
-->
antes de ayer
-->
aqui dele finish no va a ser mas nada finish
-->
no no le porte porque
-->
no vamos a crear contextos
-->
perfecto listo ahora
-->
si quiere traerse
-->
expanda por favor la conexion que creo
-->
de axa y ahi le va a aparecer
-->
las estructuras
-->
que pertenecen a esa base de datos
-->
en este caso te va a mostrar sinonimo si hay
-->
tablas y vistas
-->
los que crearon la tabla
-->
y la quieren ver para ver de este lado
-->
vamos a hacer lo siguiente aunque no la haya creado
-->
hagamos toda la misma prueba
-->
por defecto cuando yo creo la conexion
-->
no me trae las
-->
los objetos de esa base de datos
-->
me toca hacer lo siguiente
-->
conectase dele click por favor a la conexion
-->
click derecho ahora
-->
y digale donde dice
-->
extraer esquema
-->
aqui le pregunta
-->
si desea filtrar algun tipo de objetos
-->
nada mas le va a mostrar la herramienta
-->
tablas, vistas y sonodibus
-->
dele next aqui
-->
aqui expanda por favor
-->
y le van a aparecer los objetos
-->
lo que aparece vacio es porque no crearon la tabla
-->
pero entonces si deseamos
-->
exportar esa tabla dele click aqui en
-->
axa para poder
-->
seleccionar todos los objetos
-->
le da next
-->
finish
-->
y este señor
-->
ya le va a traer aqui todos los objetos
-->
que pertenecen a esa tabla
-->
ok
-->
eso funciona exactamente
-->
igual independiente del motor
-->
MySQL, Postgres, Oracle SQL Server
-->
por defecto la herramienta
-->
no te hace la conexion
-->
pero tienes que hacer de forma manual extraer
-->
los diferentes elementos que quieres de esa base de datos
-->
eso porque lo hacen de esa manera
-->
porque hay veces que base de datos tiene muchos componentes
-->
o muchos objetos
-->
y te puede demorar mucho
-->
entonces lo ideal es que crees la conexion
-->
y exportes o traiga los objetos
-->
que tu quieras manejar
-->
en la conexion como tal
-->
vamos entonces a hacer lo siguiente ahora
-->
vamos a crear aqui un tepre job
-->
aqui en el job
-->
un tepre job
-->
no, no hay problema
-->
deje la tabla, ahora cuando creemos la tabla aqui en el job
-->
hace lo mismo ahora
-->
para que le devuelva
-->
la tabla que vamos a crear en el job
-->
no hay problema
-->
esa practica la hacemos ahora cuando ejecutemos este job
-->
tengo aqui
-->
un tepre job por favor
-->
y al igual que con HDFS
-->
hay un tehy connection
-->
tehy connection
-->
conecta el tepre job
-->
con el tehy connection
-->
listo
-->
ahora vamos a configurar el tehy connection
-->
si tu quieres puedes
-->
volver a configurar todo nuevamente
-->
pero como ya tenemos la configuracion
-->
y que se encuentra en el repositorio
-->
tienes que irte a property type
-->
y le vas a decir ahora que no vas a
-->
construirla sino que vas a tomar una
-->
configuracion que esta en el repositorio
-->
le das aqui repositorio
-->
te cambia entonces
-->
aqui listo
-->
dale click ahi en el mas mas que aparece
-->
y escoge la conexion
-->
de hay que esta habilitada
-->
aqui la escojo y doy ok
-->
con esta distribucion no hay mucho
-->
problema porque al fin y al cabo
-->
quien maneja la conexion con hadut
-->
es hay
-->
y no en este caso esta en
-->
big data entonces lo dejo ahi
-->
con eso esa configuracion
-->
agrego un
-->
tepos job
-->
y voy a colocar aqui
-->
un tehy close
-->
como buena practica
-->
para cerrar la conexion porque si no
-->
puede quedar abierta y
-->
ocuparte memoria
-->
esa conexion que no lo vas a usar
-->
pero quedo abierta entonces
-->
lo configuras con la conexion
-->
que vas a cerrar
-->
perfecto listo entonces
-->
agreguemos los siguientes componentes
-->
ahora vamos agregar aqui
-->
un tehy
-->
create table
-->
y acto seguido vamos agregar aqui
-->
un
-->
tehy row
-->
tehy row
-->
y
-->
se me escapo
-->
agrego un tehy row generator
-->
listo son los componentes
-->
que vamos a utilizar entonces
-->
conecta el
-->
tehy row generator
-->
ya te explico para que sirve eso
-->
con el tehy row
-->
vas a conectar el
-->
tehy create con un
-->
sub job diferente entonces click
-->
derecho disparador
-->
con sub job ok
-->
hacia el tehy row generator
-->
ahi tendria
-->
4 sub jobs el del tehy pre job
-->
el del tehy create
-->
table el del tehy row
-->
generator y el del tehy post job
-->
vamos la primera confesion del tehy create
-->
table entonces vamos a darle doble click
-->
listo doble click a ese tehy
-->
create
-->
voy a hacer una tallita sencilla entonces
-->
en primer lugar voy a decirle que
-->
ya tengo una conexion prestalecida use
-->
conexion existente y la escojo
-->
tehy connection
-->
nombre de
-->
tabla por favor entre comillas
-->
dobles coloquele clientes
-->
si estamos le coloque
-->
clientes ahi perfecto clientes
-->
en el action on
-->
table hay varias opciones
-->
entre las opciones que aparecen nada mas hay dos
-->
una que es crear tabla
-->
el problema de crear tabla es
-->
que si la tabla esta creada ya
-->
va a salir un error
-->
entonces si este job lo voy a
-->
ejecutar ene veces lo mejor
-->
es que crea tabla si no
-->
existe
-->
para evitar es el mismo problema
-->
que teniamos con el
-->
con el del cluster de
-->
hadoop cuando creamos el archivo
-->
en el output habia que sobre escribirlo
-->
si lo ejecutaba varias veces
-->
bien y este señor le permite
-->
a usted que hacer
-->
en el formato de tabla
-->
permite hacer varias cosas
-->
o varios formatos
-->
dele click por favor en los formatos
-->
que estan prestablecidos ahi
-->
y notara que hay diferentes formatos
-->
el primer formato que aparece
-->
es el de text file
-->
text file indicaria
-->
en que los datos se van a guardar
-->
en texto plano en el
-->
por ende al estar en el
-->
texto plano ahi
-->
pues yo creeria que la unica ventaja
-->
que vas a poder hacer ahi es que es
-->
legible tu puede hacer consultas
-->
ese ql ahi y
-->
vas a entender porque esta en plano
-->
pero el problema que vas
-->
a tener con un text file es que
-->
no hay ningun tipo de compresion
-->
entonces como no hay
-->
comprension vas a ocupar
-->
mas espacio el rendimiento va a ser
-->
mas lento y evidentemente vas
-->
a consumir mucho mas espacio porque ni siquiera
-->
haces comprension de los archivos
-->
los archivos no se guardan comprimidos
-->
listo ese text
-->
file yo
-->
diria que es posible usarlo siempre
-->
cuando tu tengas
-->
mucho espacio y donde
-->
quizas el rendimiento de las consultas
-->
no requiera
-->
unos tiempos de respuesta
-->
bastante bajos porque
-->
la ventaja va a tener es que el rendimiento va a ser
-->
muy lento
-->
despues sigue ahi le aparece uno
-->
que se llama sequence files
-->
ese sequence files
-->
es propio de jadut
-->
ahora este jadut
-->
va a guardar
-->
los datos comprimidos
-->
y recordaras un poquito el dia de ayer
-->
cuando guardamos
-->
un archivo comprimido
-->
que tu lo abrias y ese archivo estaba en binario
-->
no entendias lo que ibas a hacer
-->
ahora al guardarlo
-->
de esta manera jadut lo prepara
-->
para que se ejecute tambien
-->
en forma paralela
-->
pero creo que la ventaja
-->
que vas a tener es que es
-->
dependiente de jadut
-->
significa entonces en que para
-->
poder reutilizar esto en una infra
-->
infraestructura diferente
-->
tiene que estar casada con jadut
-->
pensar en cambiar de a un synapse
-->
a un mr
-->
aunque mr usa jadut
-->
pensar de pronto en un
-->
s3 por ejemplo
-->
ahi si vos tenes inconvenientes
-->
despues sigue
-->
dos en particular que dice rcfile
-->
y orcfile
-->
basicamente son los mismos formatos
-->
la diferencia es que orcfile
-->
es mucho mas
-->
eficiente que el
-->
rcfile te preguntaras si es
-->
mas eficiente porque nada mas dejan
-->
orcfile porque el problema que tienes
-->
es que hay una incompatibilidad
-->
con esos dos formatos
-->
entonces quizas hay
-->
sistemas que ya lo tienes con rcfile
-->
y quisieras mantenerlo
-->
entonces por eso se mantiene los dos
-->
ahora una caracterÃstica
-->
que tiene rcfile
-->
que tiene
-->
uno que se llama
-->
y que creo que
-->
mis calculos no me fallan porque
-->
me parecio ver un job una vez que me reunÃ
-->
con alguien que trabajaba para ustedes
-->
es muy parecido
-->
a ese señor que se llama parket
-->
que esta ahi abajo
-->
el formato parket quizas lo has escuchado
-->
que aclaro ese formato
-->
la idea del formato no es nueva
-->
eso venia antes
-->
ese formato corresponde
-->
a un tipo de base de datos
-->
que se llama tipo columnar
-->
que favorece
-->
las operaciones
-->
que tu vayas a hacer
-->
eso por ejemplo en el caso
-->
de colombia
-->
yo te habia comentado el dia de ayer
-->
que hay una empresa
-->
que es un supermercado
-->
en que tiene varias
-->
sucursales por ciudad
-->
osea toda colombia seguramente en mexico
-->
debe haber alguna muy parecida
-->
aca se llama super tiendas olimpica
-->
entonces esos
-->
señores tienen demasiado
-->
dinero pero bastante dinero
-->
y ellos mantienen entonces
-->
varias
-->
en una misma ciudad puede haber varias
-->
sucursales aunque
-->
ellos tienen ahora una
-->
una pelea fuerte
-->
con una competencia dura
-->
que llaman aqui en colombia
-->
supermercados de bajo costo
-->
entonces hay
-->
una tendencia en que ya estan
-->
vendiendo mucho menos porque esos supermercados
-->
tambien la idea es
-->
ocupar toda colombia pero bueno depende de eso
-->
ese ors files
-->
es mas tipo columnar aqui
-->
coloco un ejemplo muy sencillo de pronto
-->
para que lo tengas en cuenta
-->
suponte que tu tengas aca
-->
sucursal
-->
columna sucursal
-->
tenga por si te
-->
algo ventas
-->
aca colocamos estas dos
-->
tengas aqui sucursal x
-->
ventas
-->
12 tal
-->
sucursal y
-->
ventas fueron 3 14
-->
z
-->
15
-->
o te lo voy a colocar
-->
para que sea mucho mas sencillo de entender
-->
voy a colocar todos
-->
en sucursal x
-->
para que me quede mejor el ejemplo
-->
si a ti te preguntaran
-->
te dijeran o
-->
te pidieran que sacaras
-->
el total de ventas que tienes
-->
al momento de la sucursal x
-->
tu podias hacerlo con
-->
ese ql o lo que uno llama con
-->
tecnologia tradicional seria
-->
un select listo puedes agrupar
-->
pero el agrupar
-->
es un sum
-->
de la columna ventas
-->
lo que pasa es que
-->
internamente lo que hace el sql
-->
es que tendria que sacar
-->
registro por registro e ir sumando
-->
yo podria mejorar
-->
esto por ejemplo haciendo
-->
particiones
-->
las particiones velo como si
-->
fuese una carpeta imaginate que yo
-->
crea una partición por sucursal
-->
imaginate que cada sucursal
-->
se guarda a nivel de una secundaria
-->
por una carpeta diferente
-->
cuando tu haces un group by por la
-->
sucursal ya el sabe que
-->
todo lo que esta en la carpeta por
-->
si necesita algo x son de la
-->
sucursal x
-->
es mas eficiente a la hora de hacer
-->
los calculos pero todavia
-->
me quedo en que me toca sacar
-->
registro por registro e ir sumando
-->
la idea esta de columnar
-->
que en realidad no es
-->
nueva es que no
-->
lo guarde en ese
-->
columna sucursal
-->
imaginate que esto lo guardo en un solo
-->
archivo sucursal
-->
sucursal
-->
y aqui guardo nada mas x
-->
y aca en ventas guardo
-->
los valores 12,
-->
14
-->
y 15
-->
que diferencia hay que cuando ahora tu vas a hacer
-->
la sumatoria de las ventas
-->
yo en realidad nada mas traigo este archivo
-->
que es la columna ventas y voy sumando
-->
efectivamente si quiera
-->
hacer un join habia que hacer
-->
un join de este archivo
-->
practicamente con quizas
-->
este archivo que seria otra columna
-->
cada columna se guarda en archivos diferentes
-->
pero la ventaja que tengo es que como solamente saco
-->
un solo registro
-->
de esa columna sumar eso
-->
y hacer operacion de matematica pues
-->
eso la diferencia
-->
con la tradicional pues
-->
es abismal
-->
a nivel de tiempo de respuestas
-->
entonces ese archivo
-->
orsl te guarda en formato
-->
formato
-->
aclaro
-->
del sequence files hacia el parquet
-->
perdon hacia el
-->
storas te guarda en binario
-->
parquet si no estoy mal te la guarda en
-->
ahora reviso para ver que ahora no recuerdo como te la guarda aqui en hide
-->
pero
-->
aclaro que parquet tambien se puede usar en otros motores
-->
no solamente de hide
-->
es un formato de archivo y que
-->
es usado por diferentes motores de baseados
-->
abro te lo guarda en binario
-->
estoy seguro
-->
input y output format
-->
ese que esta ahi al igual que
-->
el storas
-->
eso te lo coloca
-->
si tu quieres definir un
-->
formato especial de almacenamiento
-->
es si tu defines el formato
-->
no tomas un formato convencional
-->
yo particularmente nunca
-->
metido eso porque es desgastante
-->
normalmente ya estos formatos
-->
que estan ya estan ha sido
-->
testeado y funcionan bien
-->
y por ultimo tendrias el parquet
-->
que es el que te mencione que esta en colunar
-->
entonces tu tienes que definir como
-->
vas a guardar los datos en este caso en particular
-->
quiero que sea text file
-->
y esta base
-->
de datos esta tabla perdon no va a tener
-->
particiones no voy a crear particiones
-->
porque la cantidad de registros que voy a almacenar
-->
es elevadamente pequeño
-->
listo entonces
-->
eso es lo que hace ahi
-->
este row generator
-->
para que este te genera filas
-->
cuantas la que tu quieras
-->
para que sirve eso
-->
aparte de probar algunas cosas como la que voy a hacer ahora
-->
que voy a generar unos
-->
aleatoriamente voy a generar unos filas
-->
para hacerlas almacenadas en la tabla
-->
este row generator
-->
es muy bueno cuando quieres hacer pruebas de estres
-->
por ejemplo
-->
tu quieres un sistema
-->
y quisieras saber cual
-->
la eficiencia de ese sistema
-->
entonces tu haces una prueba de estres
-->
entonces empiezas a insertar una cantidad
-->
de registros y miras como se
-->
comporta con esos registros
-->
despues haces una proyeccion
-->
y esa proyeccion matematica te va a definir
-->
cuanta memoria, cuanto recursos
-->
necesitas minimo para atender a
-->
en usuarios entonces este row generator
-->
es muy bueno para hacer esas pruebas de estres
-->
vayase
-->
vamos a conferir asi el row generator
-->
por favor row generator
-->
y vamos a
-->
agregar dos columnas
-->
la primera columna
-->
la va a llamar por favor
-->
identificacion
-->
todo en minuscula
-->
y la segunda la va a llamar
-->
nombres
-->
el tipo de datos define
-->
lo string y por favor
-->
ahora en las funciones que estan aqui
-->
porque esto se viene aleatoriamente
-->
dele click en la primera funcion
-->
de el campo identificacion
-->
y va a buscar
-->
uno que se llama
-->
perdon que pena hay
-->
algo que guarde, que genere un código
-->
espere lo busco por aca
-->
que se me escapo el nombre
-->
de row generator
-->
momento por favor
-->
a listo busquelo
-->
en numeric
-->
ahi aparece numeric
-->
no se si le aparece numeric
-->
ah no aparece
-->
que raro que no aparezca
-->
perdon cambia el tipo de datos ya se porque no aparece
-->
el tipo de datos cambia al integer mejor
-->
para generar un consecutivo
-->
cambia al integer ese es el problema
-->
integer y aqui escoge
-->
uno que se llama
-->
numeric sequence
-->
que ya lo hemos usado anteriormente
-->
numeric sequence
-->
y eso te va a generar un aleatorio ahi
-->
1,2,3,4,5,6,7,8,9
-->
numeric sequence
-->
aquà he escogido que no era
-->
y aqui lo puedes confiar recuerda que esto lo hicimos
-->
en el dia de ayer cuando obtuvimos
-->
los tres primeros
-->
productos menos vendidos
-->
y en el nombre
-->
busquen uno que se llama
-->
talent data generator
-->
get first name
-->
ahi no se porque le di enter
-->
listo
-->
get first name
-->
quedaria mas o menos aqui
-->
el primero es integer
-->
numeric sequence
-->
y el segundo es de tipo string
-->
talent data generator
-->
get first name
-->
el valor numerico
-->
perdon no lo escuche el cual integer
-->
el primero
-->
es integer si
-->
no el numeric de 1,9
-->
no deje lo que lo
-->
ahi va de uno en uno pero el
-->
a partir de note lo siguiente
-->
que aqui le va a pasar numero de filas
-->
que va a generar esa 100 no la genera
-->
porque demora mucho insertando entonces vamos
-->
a generar solamente 20 y nada mas
-->
cambia ese number row para que genere
-->
20 listo
-->
ok aqui
-->
y entonces
-->
se me escapo algo en el
-->
create table que no defini el esquema
-->
osea las columnas que voy a
-->
definir entonces
-->
vaya nuevamente al create table
-->
editar esquema
-->
y agregue dos columnas
-->
una columna que se llama
-->
identificador
-->
de tipo entero
-->
y la otra columna se llama
-->
nombre de tipo string
-->
hay muchisimas gracias
-->
porque si despues no no era compatible
-->
integer muchas gracias
-->
se me habia escapado
-->
perfecto aqui esta
-->
data generator
-->
aca la llamo identificacion pero bueno
-->
no importa haga lo siguiente
-->
ahora ya para finalizar este ejemplo
-->
vallase aqui a
-->
vamos a cambiar el nombre de esta fila
-->
porque me interesa este nombre muy importante
-->
vamos entonces
-->
a cambiar esta conexion de
-->
tegro generator al tehai la vamos a llamar
-->
clientes
-->
listo entonces
-->
accedemos al tehai row
-->
le decimos que vamos a usar
-->
una conexion existente
-->
tehai connection
-->
y este
-->
tehai row
-->
funciona
-->
especificando una sentencia
-->
SQL hay dos formas
-->
de insertar datos en high
-->
una es esa forma y la otra
-->
es a traves de un cargue de un archivo
-->
en este caso voy a hacer entonces
-->
la primera forma que es
-->
insertandola con una
-->
sentencia SQL entonces
-->
aqui donde dice query por favor
-->
vamos a definir la sentencia SQL
-->
esto es sumamente
-->
engorroso la dejo
-->
por si acaso de pronto te toca hacerlo alguna vez
-->
pero quizas la otra forma
-->
es mucho mas sencilla
-->
para esto voy a hacer lo siguiente
-->
ahi donde
-->
esta el query te va a quedar de esta forma
-->
voy a copiarla aca
-->
te va a quedar comilla
-->
doble, comilla doble
-->
dentro de esas comillas doble vamos a
-->
definir la sentencia SQL con la cual
-->
yo quiero insertar valores
-->
en este caso
-->
insert int
-->
empleados
-->
abro parentesis
-->
y ahi termino la sentencia
-->
inicial del SQL
-->
quedaria insert int
-->
empleados
-->
perdon no se llama empleados se llama clientes que pena
-->
la tala la llame fue clientes
-->
no empleados
-->
la de empleados fue la anterior
-->
no porque ahora voy a pegar en
-->
en tiempo de ejecución los valores
-->
ah si perdon perdon
-->
que pena gracias
-->
aqui le voy a dar aqui
-->
un mas
-->
para concatenar el valor
-->
y aqui es muy importante como llamaste
-->
la conexion del
-->
T-Rogenerator hacia el T-Hiro
-->
con el mismo nombre y con esto sensible
-->
yo lo llame clientes
-->
quedaria entonces
-->
clientes
-->
punto nombre de la variable
-->
con que viene del T-Rogenerator
-->
yo la llame inicialmente
-->
la primera la llame identificación
-->
entonces para hacer referencia
-->
me quedaria de esta
-->
forma
-->
punto identificación
-->
y de esa manera
-->
cada vez que
-->
entre una fila voy a tomar por modo
-->
de ejecución el valor ese
-->
mas
-->
abro comillas dobles
-->
coma
-->
y aqui me toca abrir una comilla simple
-->
para que pueda quedar
-->
mas
-->
clientes
-->
punto nombres
-->
mas abro comilla
-->
comilla simple
-->
y termina asi
-->
vea que esto es sumamente engorroso
-->
pero quizas alguna vez
-->
pueda superar algún problema con eso
-->
ahi estoy creando una
-->
sentencia dinamica
-->
vamos con la comilla simple
-->
porque como es textual
-->
es texto el bar chart
-->
entonces hay que necesariamente mandarle comillas simple
-->
voy a pegarla aca
-->
asi que
-->
a ver que aquà hay todo
-->
la pego aqui
-->
cuando termines
-->
por favor manda a ejecutar
-->
el job
-->
confirma que el job
-->
que vayas a ejecutar es el que
-->
el que necesitas
-->
voy a ir de aqui a ver si no tengo
-->
algún error de
-->
bueno ahi empezo
-->
ya creo la tabla esta intentando
-->
insertar esta demorando
-->
si claro
-->
ahi esta demorando para insertar
-->
pero ya inserto la primera fila
-->
ahi esta que esta aqui
-->
si desea la coloco
-->
en la carpeta compartida
-->
Beatriz creo que salio error
-->
este para ver Beatriz
-->
para ver que error tiene en la ejecucion me parecio que vi un error
-->
dice el valor
-->
del parametro
-->
no existe ah listo dele cancel
-->
ahi un momento vayase al teja y close
-->
que es donde aparece el error
-->
y se le olvido
-->
se le olvido colocar
-->
la conexion que va a cerrar
-->
entonces cierre ahi
-->
aca en el component list le da click en el combo
-->
boss
-->
y escoge la conexion de high que va a cerrar
-->
listo intenta ejecutar nuevamente
-->
por favor
-->
ah porque tiene poca demoria
-->
como tenemos el continuo la maquina virtual
-->
y ahora estamos ejecutando con
-->
listo por eso
-->
que esta demorando
-->
empezo la conexion listo
-->
empezo a insertar va a ver si sale algun error
-->
recuerde que por lo mismo recurso que tenemos
-->
esta demorando para insertar
-->
espere para ver si pasa el dos
-->
listo ahi inserto el primero dejelo ahi ejecutando por favor
-->
quien no le funciono
-->
listo quien no le funciono
-->
vamos a ver voy contigo Fernanda
-->
que sale un error ahi vamos a ver
-->
es un error de compilacion
-->
ok
-->
vayase al code
-->
para ver que le muestre exactamente
-->
donde esta en el code
-->
ah vayase
-->
vayase al designer code al codigo
-->
ah si si listo
-->
ah todavia no cerra el lado verdad hay que serle
-->
dele
-->
click en code vamos a ver ahi esta el punto rojo
-->
que esta marcando dele click en el punto rojo
-->
exactamente el lado derecho
-->
ah dice que esa
-->
variable no existe como le llamo
-->
usted nombre o nombres
-->
entonces
-->
entra al terror generator
-->
vayase al designer nuevamente
-->
vayase al terror generator ese que esta ahi
-->
nombres
-->
listo
-->
dele en T
-->
y cambie nombre
-->
por nombres
-->
vuelve a ejecutar nuevamente a ver ah dele cancel ahi
-->
vayase al T high close que esta marcando
-->
error ahi doble click escoge ahi
-->
la conexion que va cerrada que no la configuro
-->
y vuelve a ejecutar nuevamente
-->
a ver ahora si listo ya creo esta
-->
insertando espere un momento a ver porque esta
-->
demorando mucho hay pocos recursos
-->
no deberia
-->
porque vamos a ver no no deberia vamos a
-->
ver un momento voy a revisar
-->
un momento el caso
-->
el mio para ver voy a consultar
-->
como esta mi tabla mi base de datos
-->
a ver
-->
entonces voy por aca
-->
voy a darle
-->
estoy en que en AXA no se si
-->
sele casterisco
-->
frompientes
-->
no no
-->
no dejo de insertarlo en blanco me muestras
-->
su yo para ver un momento
-->
quien fue jorge
-->
el que me dijo
-->
alejano voy contigo entonces alejano
-->
vamos a ver que paso
-->
ah
-->
te falta ahi lo que hace que te falta el
-->
punto y coma dale punto y coma ahi en favor
-->
agregale el punto y coma porque no lo
-->
no no no punto y coma porque ahi
-->
concatiendo ahi listo dale flecha
-->
ahi arriba
-->
coloco nuevamente el select lo que hace es que
-->
ahi te concateno los dos select
-->
nuevamente select casterisco entonces
-->
ahi con punto y coma exactamente enter
-->
ahi si esta manate solamente
-->
perfecto listo esa es una forma
-->
vamos entonces a la segunda
-->
eh vamos a hacer entonces
-->
el break de los
-->
20 minuticos o las 12 30
-->
hacemos los 20 minutos y estamos entonces
-->
en 20 minuticos ok
-->
listo
-->
ok
-->
ok
-->
ok
-->
ok
-->
ok
-->
ok
-->
ok
-->
Thank you.
-->
Hello again. We are back, are we connected?
-->
Yes, we are connected, or if there is audio, can you hear me?
-->
All right.
-->
Perfect, thank you very much.
-->
All right, then we create with the tool such that the only thing that does not allow or does not have a component that allows you to create database, although it is possible that you can, through an SQL pattern,
-->
I show it to you, I mentioned it on the first day, but if you want, for example, to create with the tool,
-->
sorry, I'm with Alejandro's, it's not mine, now it's mine.
-->
We can make an SQL pattern that generates the database, but hey, we won't have time to do that part.
-->
Let's look at the table, how would we do to look at the table?
-->
First of all, take into account that some of us had already created the employee table and we have created a new table.
-->
So, to be able to view the new table, we are going to do the same thing again.
-->
We select the connection to Hive.
-->
It gives you extract scheme.
-->
Next.
-->
Expand, please, there is Axa.
-->
And there you see new customers, which is the one you have.
-->
Please choose it. Next.
-->
Finish.
-->
And two tables should appear to you.
-->
To another, only one table will appear.
-->
How do we do now to process those data, for example, that I want to read with the tool?
-->
To use then tables that are defined and perform operations of some job, do the following, please.
-->
We are going to create a new job in Hive.
-->
Create job. Call it job.
-->
Read.
-->
Underscore data. Underscore table. Underscore Hive.
-->
Here finish.
-->
And very simply.
-->
It is located in the Axa Hive connection metadata.
-->
Expand the tables.
-->
Take the table that you are going to read, which in this case is customers.
-->
And release it in the canvas.
-->
By default, an ELT High Input appears.
-->
Wait a moment, if this man leaves me this component, I do not remember.
-->
Perfect.
-->
One moment, done.
-->
I'm going here to Telok Row.
-->
And it links with Telok Row, one moment here.
-->
No, it's with...
-->
Wait a moment, I do not know if I used the... Oh, I'm going, one moment.
-->
If I used the right component.
-->
No.
-->
I think...
-->
Let me confirm, because I have that ELT.
-->
And I think the component is not...
-->
ELT Input.
-->
But T-High Input.
-->
Let me check it for a moment, Telok Row.
-->
I already got to where this moment associated the problem.
-->
I'm going here to the main.
-->
I think it also works for me, I do not remember this part here.
-->
Let's see.
-->
Yes.
-->
I also test it with the other one, which is ELT Input.
-->
I think it gives me the same.
-->
Telok Row.
-->
I link with...
-->
No, it has to be with a shooter, then.
-->
No, it will not accept that component, I do not remember what it is for.
-->
The component, of course.
-->
This is one of the cases that failed, then.
-->
The tool to select is the most appropriate.
-->
The component is T-High Input, the one that will select.
-->
And then it connects it to a Telok Row.
-->
Who was telling me that I did not create the table?
-->
Who was it?
-->
I remember.
-->
Brenda, I'm going there, then.
-->
Well, let's see.
-->
Ah, but it's blocked.
-->
No, I'm seeing.
-->
How many rows did you send to insert?
-->
Twenty.
-->
Or...
-->
Wait, that takes time.
-->
Go to the left side.
-->
Where it says AXA High of the metadata.
-->
Right click.
-->
Give it to extract the scheme for a moment.
-->
Give it Next, please.
-->
Expand AXA.
-->
And there it has clients.
-->
I think that's it.
-->
Then select, please.
-->
Give it Next.
-->
And it ends with Finish.
-->
That is the table that had to be created.
-->
To confirm, go to your T-High Create Table.
-->
One moment.
-->
To T-High Create Table.
-->
And let's see the name of the table that created T-High Create Table.
-->
That one that is there.
-->
In the middle.
-->
In the canvas.
-->
That one that is there.
-->
Client, perfect.
-->
Now, do the following to confirm if the data was stored correctly.
-->
Go then to Hive.
-->
And it will create a new job there in Hive.
-->
Which is to read data.
-->
There it has it.
-->
Ready.
-->
Then take clients.
-->
And release them in the canvas.
-->
But change there to T-High Input.
-->
Add a Tello Grow.
-->
And connect clients.
-->
With Tello Grow.
-->
And proceed to execute then the job.
-->
Ready.
-->
Proceed to execute it.
-->
Here it is.
-->
There it goes.
-->
If there it has the data.
-->
What makes me weird is.
-->
There you sent 20, right?
-->
20 data.
-->
There it is.
-->
What is out of order.
-->
Yes, there are 20.
-->
Go up to read a little more.
-->
Yes, go up.
-->
Yes, there are 20.
-->
There it is.
-->
Yes.
-->
What happens is that I had not updated the connection.
-->
To see the table that was created.
-->
That's why I didn't.
-->
Yes, I had created it.
-->
But it was necessary to update.
-->
Well, let's go.
-->
Yes.
-->
Because by default the tool does not update.
-->
The changes that there are.
-->
It does not update.
-->
Of the new sockets that are in the data.
-->
You have to update.
-->
Well, let's create now.
-->
And let's see the other way.
-->
Of how we would do to insert data.
-->
In a table.
-->
For that, then.
-->
Let's create another job.
-->
Please, if we are talking about hype.
-->
And this man.
-->
I'm going to call him.
-->
Job.
-->
Under score sales.
-->
Sales.
-->
Products.
-->
By genre.
-->
Sales.
-->
Products.
-->
By genre.
-->
Let's go then.
-->
To mix a bit of.
-->
Big data.
-->
Of progress.
-->
With storage in hype.
-->
At the end it gives you finish.
-->
Then.
-->
Of hype.
-->
You can copy.
-->
In the previous hype.
-->
Copy the tepryok that serves me.
-->
The tepryok.
-->
Copy.
-->
Paste it here.
-->
And look for a.
-->
In the.
-->
The hadut.
-->
Look for a.
-->
TH.
-->
HDFS connection.
-->
To paste it there too.
-->
And reuse it.
-->
This HDFS connection.
-->
OK.
-->
Copy.
-->
And paste it here.
-->
Perfect.
-->
And I connect it.
-->
In the same tepryok.
-->
With the tehy connection.
-->
I copy the tepos.
-->
I close the connection.
-->
Of the hype.
-->
That you have it here.
-->
Ready.
-->
This would be the original structure.
-->
Then.
-->
Copy.
-->
And I'm going to paste it here.
-->
In.
-->
Perfect.
-->
This one here.
-->
And this one here.
-->
Those are the elements that I need now.
-->
The what?
-->
The tehy connection.
-->
Yes.
-->
I had had that.
-->
If you don't have it.
-->
Add it please.
-->
We had done that in the previous one.
-->
Add a tepos.
-->
And a tehyclose.
-->
Remember that tehyclose.
-->
Will show the connections that the job has.
-->
And you have to specify here.
-->
The connection that is going to close.
-->
Otherwise it will be a compilation error.
-->
That is not the past.
-->
It is the same because it takes the configuration.
-->
Maybe you ask yourself.
-->
You start to ask yourself.
-->
There is no way that I.
-->
Don't always have to copy and paste this.
-->
The same.
-->
Well, in the version of the Talent Studio.
-->
The commercial.
-->
They have what is called a Joplex.
-->
A Joplex is a set of components.
-->
That is involved in a job.
-->
So I paste that Joplex.
-->
And I could have a Joplex.
-->
That I have the pre-job and the tepos-job.
-->
So there.
-->
The only thing I do is paste the job.
-->
And I don't have to copy and paste.
-->
Otherwise I already have everything in one part.
-->
And I can reuse it many times.
-->
I hope I have time to show you a little bit of that part.
-->
Well then.
-->
I need the following components.
-->
I need an HDFS input.
-->
Input.
-->
I need a new component.
-->
Here a T-Aggregator Row.
-->
I'm going to do the following.
-->
Here a T-High Create Table.
-->
No lie.
-->
It makes more sense.
-->
The T-High.
-->
Makes more sense here in the T-Pre-Job.
-->
The Create Table.
-->
It makes much more sense.
-->
I'm going to put it down.
-->
I'm going to connect it here.
-->
On Component OK.
-->
It makes more sense.
-->
In the T-Pre-Job.
-->
And I'm going to finish here.
-->
To upload these data.
-->
An HDFS Output.
-->
And I'm going to finish with a component.
-->
That is called T-High Low At.
-->
It would be the components that I'm going to have here.
-->
HDFS Input.
-->
T-Aggregator Row.
-->
T-HDFS Output.
-->
Perfect.
-->
Ready.
-->
Well.
-->
Now then.
-->
We are going to configure the T-High Create Table first.
-->
We are going to use an existing connection.
-->
And the name of the table.
-->
We are going to call it.
-->
We are going to call it.
-->
Under score by.
-->
Under score genre.
-->
Sales.
-->
Under score by.
-->
Under score genre.
-->
In the Action Table.
-->
I'm going to tell you then.
-->
That create the table if it doesn't exist.
-->
If you are going to get a little high.
-->
You have to investigate a little.
-->
The partitions if necessary.
-->
Create partitions.
-->
But the interesting thing is that the tool allows you to configure that.
-->
And the.
-->
Scheme.
-->
I'm going to take it from what they were down here.
-->
So.
-->
So as not to do it twice.
-->
I configure it down here.
-->
And then I export it.
-->
And I import it up here.
-->
I'm going to do the following now.
-->
Let's go.
-->
And we connect.
-->
To the HDImpute.
-->
We connect it to the T-Aggregator Row.
-->
And the T-Aggregator Row.
-->
We connect it to the HDFS Output.
-->
Perfect.
-->
There.
-->
Double click.
-->
In the top part.
-->
Double click.
-->
Up there.
-->
In the header.
-->
I'll tell you where.
-->
For the moment.
-->
If you double click here in the header.
-->
He's going to put you.
-->
Give him another double click.
-->
I'm not going to leave.
-->
That you lost the.
-->
You lost this one here.
-->
This is the one that remains.
-->
Double click.
-->
Again.
-->
Up here.
-->
Ready.
-->
So.
-->
Let's go to the HDImpute.
-->
To configure it.
-->
We say to use existing connection.
-->
HDFS Connection.
-->
Take as a reference.
-->
The file that is called.
-->
Let's connect here.
-->
That is called.
-->
Supermarket.
-->
Sales.
-->
And I'm going to check that supermarket.
-->
If it's separated by what.
-->
If by comma.
-->
Or by dot com.
-->
But if by comma.
-->
By dot com.
-->
I'm going to check.
-->
To see what I forgot.
-->
Separated by comma.
-->
And they have header.
-->
Ok.
-->
So.
-->
I tell you in the input.
-->
I tell you.
-->
Well.
-->
Text file.
-->
Separator.
-->
Is.
-->
Comma.
-->
And the header is.
-->
One.
-->
Then.
-->
I'm going to look for the scheme.
-->
That surely I have it there.
-->
Temporal.
-->
I'm going to import it.
-->
Temporal.
-->
It must have a supermarket.
-->
Out there.
-->
I think we had read it already.
-->
This supermarket.
-->
XML.
-->
Perfect.
-->
Ready.
-->
Let's go to the aggregator row.
-->
Well.
-->
Since I want the sales of products by genre.
-->
Then.
-->
Let's go to the aggregator row.
-->
Let's edit the scheme.
-->
And let's go then.
-->
Gender to the right side.
-->
And product line.
-->
I do this in this way.
-->
It is to keep the names.
-->
So.
-->
It makes it easier for me to take the names.
-->
And the same type.
-->
Than adding the new ones.
-->
But.
-->
If I want.
-->
I can add it here with the plus.
-->
And let's go there.
-->
Total also.
-->
For this side.
-->
I do it just for that.
-->
To keep the names.
-->
The same names of output in the file.
-->
It makes it much easier for me.
-->
Ok.
-->
Then we group by.
-->
Gender.
-->
And by product line.
-->
Yes.
-->
Yes.
-->
Of course.
-->
Yes.
-->
Of course.
-->
Fernanda Lisbe.
-->
I already connected.
-->
Let's see.
-->
Go to.
-->
Edit the HFC input scheme.
-->
If you are kind.
-->
There where it is.
-->
Go to import.
-->
And surely.
-->
In the temporary folder.
-->
It has to have.
-->
An XML of supermarket.
-->
The last one it has.
-->
Perfect.
-->
Ok.
-->
In the separator then.
-->
Of field.
-->
It is a comma.
-->
Not point and comma.
-->
And the header.
-->
It is one.
-->
Because it has a header.
-->
Perfect.
-->
Let's go then.
-->
To the t-aggregator room.
-->
And delete the scheme.
-->
Go to the right side.
-->
Because it makes it easier for me to configure.
-->
To keep the same names.
-->
Go to gender.
-->
Product line.
-->
To the right.
-->
Ready.
-->
Product line.
-->
And total.
-->
Which is a little lower.
-->
Ok.
-->
Ready.
-->
Now let's go to the following.
-->
Let's define the operation.
-->
The operation as it is total.
-->
Give it a click here.
-->
And the total change appears.
-->
And the function.
-->
Is sum.
-->
Ignore the nulls.
-->
Because I don't remember if there are nulls.
-->
Yes.
-->
Please.
-->
Ignore the nulls.
-->
And there you don't have to do anything else.
-->
Let's go to the HF output.
-->
Let's say that we are going to use an existing connection.
-->
Let's go here to HF connection.
-->
And ready.
-->
And we are going to choose a file.
-->
To later change the name.
-->
So I'm going to call it.
-->
Sales.
-->
Under score.
-->
By.
-->
Under score.
-->
Gender.
-->
Well, here I am progressing with Big Data.
-->
In which.
-->
Process.
-->
The data that I have.
-->
And I send it an output file.
-->
To the Hadoop Cluster.
-->
Initially what I did was that.
-->
Remember that the Create Table.
-->
We didn't manage to configure the.
-->
The.
-->
Scheme.
-->
Because the scheme depends on the structure of the output file.
-->
So.
-->
Since I have the structure of the output file.
-->
I'm going to edit the scheme in the.
-->
In the output.
-->
And export the scheme you have there.
-->
Call it.
-->
Sales.
-->
Under score.
-->
By.
-->
Gender.
-->
XML.
-->
You always have to use the XML.
-->
Because then you don't read it.
-->
Ok.
-->
I export it initially.
-->
And now I'm going to import it in the Create Table.
-->
So the table will remain.
-->
With the same scheme.
-->
Import.
-->
And sales by gender.
-->
And there I have the scheme.
-->
What I don't like about how it turned out.
-->
Was that it has the first in capital letters.
-->
I don't know.
-->
As usual.
-->
I'm going to leave it like that in the middle.
-->
Well.
-->
Let's continue then.
-->
Let's go here.
-->
Where it says.
-->
Well.
-->
We are going to.
-->
That the subjobs are different.
-->
I'm going to connect then.
-->
The subjobs of the HDF input.
-->
With a subjobs of the Teja and Loa.
-->
Remember.
-->
To do the subjobs.
-->
The first component of the subjobs.
-->
Connects it with the first component of the second subjobs.
-->
Keep in mind.
-->
That here.
-->
I don't know if I put it.
-->
Because by default.
-->
The high index file.
-->
Wait comma.
-->
It escaped me.
-->
So.
-->
In the HDF opus.
-->
Please.
-->
Place comma.
-->
As separator.
-->
And it escaped me.
-->
That includes headers.
-->
Perfect.
-->
It's there.
-->
We are almost there.
-->
Perfect.
-->
So.
-->
Let's configure.
-->
The T.
-->
In the output.
-->
Of comma.
-->
Perfect.
-->
No problem.
-->
Well.
-->
Here.
-->
As suddenly.
-->
In the opus.
-->
We can execute it several times.
-->
Please.
-->
Place it there.
-->
Over white.
-->
Here.
-->
As a curious thing.
-->
If the guy.
-->
In the opus.
-->
If the guy.
-->
Had chosen.
-->
As sequence files.
-->
The table.
-->
It has to be of sequence type.
-->
Because.
-->
Otherwise.
-->
They are not compatible.
-->
Let's go to Teja and Loa.
-->
Let's tell it.
-->
Use existing connection.
-->
Perfect.
-->
There.
-->
This Teja and Loa.
-->
It has two ways to load.
-->
If you go in the Loa action.
-->
The Loa appears.
-->
And if the second option.
-->
Choose insert.
-->
Practically.
-->
It comes out.
-->
The same menu.
-->
That I had now.
-->
With the Teja and Loa.
-->
So.
-->
I'm interested in.
-->
How to do it with the Loa.
-->
This Loa.
-->
It has to be much faster.
-->
Because what it's going to load.
-->
Is a file.
-->
Directly.
-->
It's not going to be a SQL sentence.
-->
If not.
-->
It's going to be a bulk.
-->
To load.
-->
The records.
-->
To the table.
-->
So.
-->
Let's go here.
-->
The PAD.
-->
The PAD.
-->
Look at the following.
-->
Down here.
-->
There's a local.
-->
If I click this local.
-->
It's going to look for the file.
-->
That references here.
-->
In the location.
-->
That is on this side.
-->
That is to say.
-->
That if the local.
-->
I leave it here.
-->
In reality.
-->
It would be looking.
-->
The local.
-->
The file.
-->
That I place here.
-->
In the container.
-->
But I want the file.
-->
To be in Hadoop.
-->
So.
-->
I'm going to click this local.
-->
That is here.
-->
I am going to copy.
-->
The file.
-->
The route would be.
-->
This output.
-->
Is here.
-->
I'm going to copy the route.
-->
Output.
-->
That will be the file.
-->
For the Lua.
-->
Copy.
-->
And I paste it.
-->
In the Tejai.
-->
Lua.
-->
In the Failpad.
-->
In the Failpad.
-->
Of the Tejai.
-->
Lua.
-->
And the table.
-->
Let's call it.
-->
I'm going to copy the name.
-->
For.
-->
no problem, I copy the name and paste it here, just in case I have to save it in case of doubt,
-->
who takes it? Ok, ok, well, I'm going to try, I think everything is fine here, I'm going to try to
-->
run to see if there is no error, this should create the table and load the data that is
-->
in the output output file, I'm going to give it a run here to see what comes out,
-->
it is starting the process of bin data with my reduce, it created and supposedly there is no error,
-->
I'm going to see if it created the table, there are some that came out in error, not me, I'm going to
-->
see my first review here with me, I'm going to see, I'm going to extract them again,
-->
here it is, I think the table is sold, perfect, I'm going to consult it quickly here,
-->
if there is no error, there is no error, select asterisk from sales
-->
or under the core genero
-->
and the problem is that it inserted it to me but in a single column, wait a moment to see,
-->
I have run the errors, I have an error, I think with the, with the comma, let it load,
-->
ah no, what happens is that yes, yes, here I had a problem because the escape character of the
-->
table by default was in point and comma, change the escape character of the table please to
-->
escape this part and there was the problem, I'm going to try to run it again
-->
and there should be a problem with the separation characters, I thought that by default it was
-->
comma and in reality it is point and comma, I see that there are some that have errors, I'm going to
-->
ready, apparently ready, let's see
-->
no, I got the same error, ah, no, here is a problem, the problem I have is that, let's see,
-->
what is the problem I have now, the problem is that the table is already created, I will not
-->
create it again and the table has by default, I had, sorry, by default the separator point and comma,
-->
so it is easier for me because it is already created, I will not create it again, it is easier for me to
-->
change here in the output so that it is point and comma, the separator, I thought the separator was
-->
by default comma, then point and comma, let's see now yes, run again,
-->
I see that you have an error there, I do not know if you have overwrite in the output,
-->
you check to see if you have it overwrite and I try to run again and now yes, if the problem was
-->
that it was already created then the separator is point and comma, let's see, here it did not work,
-->
notice that the response time was superior, well, what did I do, I don't know, what a shame,
-->
I interrupted you, I'm going with you for a moment then, well, let's see what it says there, let's see,
-->
in the third line where it says Java file no phone, it did not find the file, it seems to me that
-->
go to hfimp if the configuration is wrong, of course, the problem is that our next one,
-->
I would be looking for that file in the root of the Hadoop closet, exactly, go there,
-->
if you want, but wait a moment, that is for the output or input, a moment, a moment,
-->
I left, I left, let me connect in an iterative way to be able to, a moment please,
-->
close the window for a moment that I left, a moment, close the window please,
-->
no, that input has to be, it is supermarket, then go back to the input, there you have it,
-->
sorry, go back to the, there, perfect, go to AXA and there it is supermarket,
-->
exactly, it is the entrance, ok, the output, please, the output, yes, that output is fine,
-->
let's go to Teja and Loa to see, to confirm, and it's fine, run to see.
-->
Well, there came out a small warning, I don't know, it already exists, well, it's over, if you want,
-->
do the following, let's do all the following to confirm the data then,
-->
that I did it over there, it was through the console, ready, let's do then, read data,
-->
ok, those who have already finished, please add a new job to read the table,
-->
remember that you have to refresh the scheme and then make an input and a telorow, ready,
-->
let's see, let's go to the output, please, to the output, ready, that's fine, let's go to
-->
Teja and Loa, remove the local, please, remove the local because it was looking for it there,
-->
the local chulito, because it was looking for it by default in the container and it is not there,
-->
try to execute again to see, ready, it's over, then, let's leave it there still,
-->
that we are going to do, we are going to consult it with the tool, to consult it with the tool,
-->
first please update the connection to the data base of Hi, how would it be,
-->
I go here to the Axa and update, extract scheme, next, next, Axa here, and here it was already,
-->
we selected 20 geniuses, if you don't have it, please select it,
-->
ah ok, let's see, let's see, it can't recognize, table not phone,
-->
I think it says, a little more to the right, yes, exactly, table not phone, sales by genre,
-->
ready, go to the Teja and create table for a moment, Teja and create table, copy in case that
-->
name comes from there, directly copy it, and go to Teja and load it, paste it there then,
-->
try to execute again to see,
-->
no there it says, table not phone, sales...
-->
ah, there, there, there, there, surely that is it, go to Teja and create table,
-->
And now, try to execute, yes, what happened, I tried to create the table, but I couldn't because the scheme was missing, and when I referenced it, it was created.
-->
Perfect, now do the next job, a new job in Hive that we are going to read, yes please, do that, it will be able to extract the scheme.
-->
Next, expand there AXA, choose sales in Hive, I already created it like this, next, finish, ready, then I was already in that part, create now a new job called job, read, job, read, sales table, nothing else, sales under core Hive.
-->
Here, finish, ready.
-->
Well, and then how do you do the sales table, I release it on this side, and remember that this is high input, I place a teloc row, and I send it to run.
-->
Well, here it is clear that with the tables created, and after you progress with Big Data, then what comes is pure SQL, or any component that you want to use, here it is, perfect, there it is.
-->
If anyone does not show data, please let me know.
-->
I do not know Alejandro, is there an error? Oh no, but it is Alejandro, sorry, what comes, what does it tell me?
-->
There was an error, right? Oh no, let's read the truth, then let's go to the next one, first delete this high input 2, because now, good question, because then it escaped me.
-->
First, update the connection you have of the Hive metadata, how will it be? Go to the connection and let it extract the scheme.
-->
Extract the scheme, I'll do it, extract the scheme, next, expand there please, next, finish.
-->
Now, take the sales table, expand tables, there are tables, there it expands, and sales, release it here.
-->
Now, the way you were going to do it, it also came out, but this is faster, why? Because here it already has the table, it already has the configuration of the connection to the Hive, you do not have to do it again.
-->
Give it a telocrow, give it a telocrow and send it to run, then.
-->
It executed the previous one, you have to run reading data, Alejandro, are you running the previous one?
-->
Open now, read data, no, it executed the previous one too, close and open again, close that job.
-->
Save it, sorry, save it, the engineer applies, close and open again, yes, perfect.
-->
Ready, then, we made the example with Hadut, we made the example with Hive, and actually the example with Sparse, this Open Source tool does not allow me,
-->
but I'm going to show you how easy it is to configure it in the proprietary tool.
-->
In the proprietary tool, I'm going to close here for a moment, notice that the jobs you have have the same structure,
-->
Basic Room, Debut, Advanced, Target and Memory Room, right? The jobs you create.
-->
Now, the jobs you create in the talent big data proprietary tool, this one that is here,
-->
is now going to add a new component called Spark Configuration.
-->
That Spark Configuration tool in the first place, you define where your Spark cluster is going to be.
-->
You configure it initially, and now, wait a moment, there you configure it.
-->
Spark Job Moment Talent, ready, I left it to see if I can find the graphic.
-->
Let's see, I can't find a graphic.
-->
Well, there you configure it, sorry, this configuration is done immediately in the job,
-->
and what the job is going to do is generate all the code you have, the job of the components that you have or that you have configured,
-->
it sends them and converts them to Spark in Java, and sends them to run against the cluster that you have defined.
-->
Why the example doesn't say it? Because the tool doesn't allow it.
-->
But in this case, it is connecting to the AWS MR cluster,
-->
and what it does then is that, here, let's see if it shows me the job.
-->
It shows here the job, this job that is here, which is a conventional job,
-->
the only thing we are going to configure is that you are going to tell it that you are going to run it against Spark,
-->
and it will automatically generate the code and it will make the Spark there.
-->
That's the only thing it does, it's actually very simple.
-->
You don't worry about Spark, you don't worry about the job, you do it conventionally, and in the end it sends it to run against the cluster.
-->
That would be the way to run it against Spark,
-->
I apologize because the tool doesn't allow me that part.
-->
And as a new thing, within the latest versions that the talent has,
-->
it already comes with components from machine learning.
-->
So it also saves you a lot from the machine learning processes,
-->
that normally one gets married with Python,
-->
but normally one gets married with Python because the Python ecosystem for machine learning
-->
is much larger than the machine learning ecosystem that it has for Java.
-->
It doesn't mean that Java doesn't exist,
-->
but rather that the machine learning in Python is much more mature.
-->
So to not start projects from scratch, one does it with Python.
-->
But now here, in the new versions that the talent has,
-->
it adds some components to be able to analyze the data
-->
through machine learning and with this it generates some descriptive or predictive patterns,
-->
as you want to do it.
-->
So it saves you a lot of work.
-->
These components, I don't understand much about the license part,
-->
I understand that the licenses are handled in the following way.
-->
You buy the talent studio and in the end,
-->
if you want a big data component or a machine learning component,
-->
what you do in the end is that you pay talent
-->
to enable you in the repository to update the components
-->
in the license that you have.
-->
That is the way it is working.
-->
Allow me to give an example with the version of the talent studio
-->
that does not come with support to enable that talent studio.
-->
But I'm going to show you some different things.
-->
Allow me, now I'm going to share my screen.
-->
The one of my team, not the one of...
-->
Give me a moment, this is my team.
-->
Wait a moment, ready.
-->
Here I have downloads, here I have talent studio.
-->
Ready, this is the commercial version.
-->
Wait for me, it's opening on the other screen.
-->
I'll try to pass it here.
-->
Look at the following.
-->
First, what does it start to change?
-->
It starts to change in the way the tool works.
-->
In the sense that the Open Source tool is designed more for personal projects,
-->
but I could adapt it to group projects.
-->
And this is designed for group projects.
-->
So, in the Open Source tool, I connected to a local repository,
-->
that is, a folder where I was going to save the different jobs that I was doing.
-->
In this case, what this tool is going to do is that you have to define a project.
-->
By defining a project, you associate that project with a Git repository.
-->
I don't know, they got married with Git.
-->
Although Git is the most used repository by Git developers,
-->
but I don't know if that was the decision of why they got Git at the end.
-->
But you connect to a repository in which all the people who belong to the same project
-->
are going to connect to the same repository.
-->
And that is the way they are going to share the different jobs.
-->
Here, by default, this part is already configured.
-->
So, here I say Open.
-->
Here, this version is a hybrid version,
-->
in which the authentication is done by the TMS that is in the cloud.
-->
How do I do it?
-->
In the TMS, I generate a key,
-->
and that key is the one that will allow me to log in
-->
and be able to open this talent studio application at a local level.
-->
So, how do you manage the versions of this man?
-->
Well, he manages the following way.
-->
Suppose you have a project.
-->
Well, here it tells me to wait five days.
-->
I'm going to create ... here are two types of jobs now.
-->
A job that is called Standard and a Job Lab.
-->
I'm not showing it here,
-->
for the question they asked me about how they manage the versions.
-->
Although I have it ...
-->
I could configure something in the TMS and we can wait there.
-->
So, I have little left for the evaluation that has to be done in the end.
-->
Here are two types of, let's say, jobs.
-->
One called Job and one Job Lab.
-->
The Job Lab is to reuse components within a job.
-->
So, imagine where I can use a Job Lab.
-->
Note that I, in most of the jobs I did,
-->
I always copied a pre-job and a post-job.
-->
With some basic elements that I was going to need.
-->
So, I can make a Job Lab.
-->
That Job Lab already has the pre-job and the post-job.
-->
And what I do is add that Job Lab to a particular job.
-->
In short, that Job Lab is what makes it have a set of components
-->
that prevents me from copying and pasting.
-->
It automatically makes me copy and paste.
-->
That's what a Job Lab does in the end.
-->
In the Standard, look at the interface, it is very similar.
-->
Well, graphically, there is a little thing at the level of shape,
-->
but in the end it is the same.
-->
Let's go here to Standard.
-->
Let's create a Standard Job.
-->
Let's put Job Example here.
-->
It opens the same bed.
-->
Very, very similar.
-->
Ready.
-->
So, I'm going to create a metadata here.
-->
I hope it's good.
-->
I already have a metadata here.
-->
I'm going to take this metadata.
-->
I'm going to take this Free file.
-->
The same idea.
-->
I'm going to add the Teglo Roam here.
-->
And here I connect this man with this man here.
-->
So, ok.
-->
The metadata concept will be achieved in all the talents, first of all.
-->
Note that here there may be some metadata elements
-->
that are in the OpenBit data that are not here.
-->
For example, here Hadoop does not appear.
-->
Not there.
-->
Why does Hadoop not appear?
-->
The Hadoop file that I have does not have that part.
-->
But when you buy or acquire the Big Data license,
-->
Hadoop will appear.
-->
There are the different elements that are there.
-->
In other words, in the end it is a matter of license
-->
that they want to sell their product.
-->
The question is that you buy more in a few words.
-->
So, look at the following.
-->
Imagine that here I have this Job that is here.
-->
Ready.
-->
Ready.
-->
That Job is there.
-->
I'm going to try something new.
-->
And I'm going to create what one calls a ...
-->
What is it called?
-->
When I make a partition in a KIP, I make a different branch.
-->
I make a fork in the branch.
-->
I make a fork in the branch.
-->
But I'm going to try something, but I don't want to leave the original.
-->
So I go to this man and I tell him here.
-->
Open in another version.
-->
I don't know.
-->
Here I open in another version.
-->
And here is where he charges life.
-->
The version that I'm going to read.
-->
The Major, the Minor.
-->
I'm going to create a new version.
-->
I'm going to upload the new version to 0.2.
-->
I'm going to finish it here.
-->
And I start where I left in version 1.
-->
So this version we are going to assume that I see the change.
-->
And I'm going to put a theme here.
-->
I put the theme here.
-->
I connect the theme here.
-->
Out here.
-->
Ready.
-->
I'm going to give it here and save.
-->
And now I have two versions.
-->
So if I'm wrong, I can open another version.
-->
And I can open the one here.
-->
See how the one was left here.
-->
And I can open the last one, which would be the two.
-->
And I can start from the version that I want.
-->
So that's how the versions work.
-->
So these versions are saved locally.
-->
But in the end you would have to do a push.
-->
A push that you do it automatically here.
-->
You put the comic manager here.
-->
The comic message, sorry.
-->
Here is the push.
-->
And you can see that when the other members of your group do a pull.
-->
Then what that's going to do is take the last changes you have.
-->
Obviously all the good practices that you have at the level of software development are applied.
-->
And all the good practices that you have at the level of using version control.
-->
But here it makes it very simple.
-->
Here, for example, the documentation that is generating you.
-->
For example, I tell you here add documentation.
-->
I think here, sorry.
-->
I go here and I tell you.
-->
Generate html doc, html documentation.
-->
Finish here.
-->
I think here it generates it in the folder.
-->
And this in the end what it does is that it generates some files that describe me.
-->
Well, I'm going to do it here.
-->
Generate all the documentation jobs for a moment there.
-->
It depends on what you want.
-->
I mean, what functionality?
-->
No, by default that comes already.
-->
No, that comes integrated into the talent studio.
-->
In any version it is.
-->
The only thing you have is that it is your power to define and clear the repository where your jobs are going to be.
-->
Particularly because I am married to Github.
-->
Remember that Github is a Git service.
-->
So I get married to Github and there I create my projects and I relate them when I'm working in a group.
-->
That's what generates all the documentation for this part.
-->
I don't know if.
-->
Yes, tell me.
-->
No, it can be public and private.
-->
Just let me show you something here.
-->
Where do you do that?
-->
Remember that I told you that in the end the tool is integrated, it is hybrid and it is going to integrate with a tool that is the TMC.
-->
In the TMC you are going to create as many projects as you have to develop.
-->
So every time you have a new project, in the end what you do is that it comes here.
-->
I'll show you a little bit of the TMC.
-->
This was what we used for the talent data preparation and for the talent data storage.
-->
But we didn't see the talent more console.
-->
Wait a moment and load.
-->
On the other hand, while it loads here.
-->
Here, if the good practice at the level of documentation is important because you have a tool called the talent data catalog.
-->
So in the future, to do the part of the life cycle or lineage of the data, that is, where the data starts, what changes there are.
-->
Normally, I could do it with the power center and several ETVs.
-->
But since I bought the talent studio, I would believe that it will integrate its talent studio, the jobs, with the talent data catalog.
-->
Therefore, it is vital that the names of variables are well documented.
-->
Any documentation that can, in such a way that when someone recovers it already in the talent data catalog, they can understand what is done internally in the job.
-->
Well, here I go to the TMC, which is this one here.
-->
Let's see in project.
-->
Here I have the test project that I am connected and I tell you the following.
-->
This is the URL of the repository that I am connected.
-->
So what you're going to do here is just create a new project with the name you want and here you put the URL of the repository here.
-->
And that's it, there you integrate them.
-->
And it does all the part of the push, the pull, it will do it automatically.
-->
Well, that's it, here are many more things, but hey, it doesn't give time.
-->
I don't know if we start with the part of the case study.
-->
Here I have two case studies actually.
-->
There is a case study that it seems to me that it will not give us time, we have practically two hours.
-->
I think I'll wait a moment and show you the case study that I have prepared.
-->
I think we have to change it because it will not give us time.
-->
Either we cut it or I make one that is simpler.
-->
Wait a moment and I show you the case study here.
-->
I think I lost the case study.
-->
I don't know if it's going to take a little break of 10 minutes, but it will allow me to tell you that my son is calling me.
-->
It is not customary for him to call me, so I want to answer him in case he says something that is happening.
-->
Give me 10 minutes and I'll come then.
-->
Well, I'm back.
-->
Fortunately, everything is fine.
-->
I don't know if you are listening, are you there?
-->
Yes, well, that's it.
-->
I had shared a case study to solve here in class, but it didn't give time.
-->
This is the case study.
-->
Let's read it for a moment to see if you suddenly think it is very complicated because of the time we have, practically only two hours.
-->
It says the following.
-->
The idea is to simulate a kind of data lake, the simulation.
-->
So the description says the following.
-->
As the explosive growth of music streaming platforms such as Spotify and Apple Music, a large volume of data is generated associated with the profiles of bands and musical artists.
-->
Millions of users interact daily with these profiles, waiting for recommendations and searches in real time.
-->
To support this demand, the Rockbrand music platform wanted to migrate the 12 billion profiles of artists currently managing a system of scalable and high-performance rigdata.
-->
To carry out initial tests in its traditional infrastructure, the response times for simple consultations took more than three hours.
-->
Remember that in the end, Big Data is for cost analysis.
-->
I wanted to give you this time at a maximum of five minutes to go to the users.
-->
So here is placed an initial configuration in which you are going to create a folder in the Big Data tab called case study and we are going to create three subfolders.
-->
This is a folder that resembles the three folders that you handle in the data lake.
-->
Which is the bronze folder, silver and gold.
-->
Here it gives you recommendations to use peyotips for each of the jobs you are going to do.
-->
And here are the different activities that you have to do.
-->
First you are going to download the S3 files.
-->
Then you are going to put them in the Hadoop's host, and then you are going to resolve these jobs that are here.
-->
For the time being, it will not give you time.
-->
There are eight activities that are here and this one is prepared for practically four hours and we have half.
-->
So I propose for a matter of time, let's make one that is much simpler than the one I was going to do in class.
-->
If you want, we do it together and present it, I don't know if we all do it and present it then at the end.
-->
Or if you want, we do it separately, I don't know what you think.
-->
It is this, it is to make a data warehouse in Hype.
-->
This is much easier to solve.
-->
What do you prefer? Do we all do it or do we do it individually?
-->
Let's do it together then, perfect.
-->
For this, please, what are we going to do?
-->
Create a document there in the browser, I don't know if it's Word.
-->
Open Word there so that it resolves each point and at each point it will add a screen.
-->
Ok, that's the first thing you have to do.
-->
No, the tone machine has Ophi installed, then in its machine, in its machine and there it is taking screens.
-->
Ready, I'm going to locate myself in the machine, ready, click, close all.
-->
So there.
-->
Well, the first thing then is we are going to create a folder.
-->
Ok, I think we already have that file, please confirm it if we don't do it.
-->
Confirm within the files that you have in the Hadoop Cluster if you have one called team underscore nba.
-->
Ok, this one that is here, I am going to confirm to see.
-->
Confirm please if you have it.
-->
If you don't have it then to avoid, I'm going to do the following, I'm going to upload that file in the shared folder.
-->
Wait a moment please, I download it because it is in AWS.
-->
So that you have it and we can start from there, AWS.
-->
I'm going to upload it, it's called team nba, this one that is here.
-->
Look, verify.
-->
It doesn't have it, so let's do the following.
-->
I'm going to download, I'm going to look for it for a moment and I'm going to upload it.
-->
Well, I just put it in the shared folder.
-->
And who was it that didn't have it? Fernanda was it? Was it Fernanda?
-->
Ok.
-->
So download it initially from the shared folder.
-->
The others please wait for me for a moment.
-->
Wait a moment please while I review here.
-->
Perfect, I hope I lost the share window.
-->
Wait a moment please.
-->
Ok, now yes.
-->
Ok, did you download it? Yes, you downloaded it.
-->
Perfect, let's see if you have it in the browser, in the Hadoop cluster in the interface.
-->
And look, there is a button that says upload file.
-->
Look if you can upload it so you don't have to do the job.
-->
Do it manually then.
-->
Confirm if you uploaded it please.
-->
Ok, perfect, I uploaded it right? Perfect, I did it.
-->
Now yes, let's go then.
-->
First point, let's create a subfolder called case studio.
-->
Case studio, folder case underscore studio.
-->
And remember that at the level of what is data warehouse, a set of dimensions is defined.
-->
The dimensions can be in stars or in a snowflake.
-->
We are going to make then two dimensions, practically two tables in stars.
-->
For this, they will be mounted in HAI.
-->
We are going to create then the first case studio.
-->
If anyone stays please tell me.
-->
The first case studio, we are going to call it job dimension players.
-->
Job dimension players.
-->
And copy the pre-job and post-job.
-->
The last job we did, we had the one from HAI with which we created the table, the last table.
-->
I'm going to look for it here, which would be this one.
-->
Job sell product by genre.
-->
And we copy the pre-job and post-job.
-->
We copy this pre-job.
-->
And we copy this post-job.
-->
Copy list.
-->
It is asked that the names of the dimensions are in Spanish.
-->
Then in the THAI create table, we are going to create a table called players.
-->
Players.
-->
And add a scheme of three columns.
-->
Erase the columns that I had previously there.
-->
And add the following columns.
-->
We are going to put a full type identifier.
-->
Identifier.
-->
We are going to put a full type.
-->
Add a new column called name.
-->
This is going to be of type stream.
-->
Add a new column called team.
-->
And it will be of type stream, only three.
-->
We are going to give it here, ok.
-->
And then we go to the following.
-->
Now yes, we are going to make an HDFS input.
-->
We are going to make a theme that we are going to map.
-->
The table has all the data of team.
-->
I have not explained.
-->
Team has the data of all the games that the NBA has in a year.
-->
I do not remember if it was 2010 or 2012.
-->
You will notice that the players will appear NBS.
-->
Then we are going to do the following.
-->
We are going to add here a new component that we had not seen.
-->
That is called Tejunic row to eliminate duplicates.
-->
Teunic with Q at the end, row.
-->
This is going to send to an HDFS output.
-->
And it ends up in a HLWA to load the data in the table.
-->
Well, let's connect then.
-->
Connect the HF input to the theme.
-->
The theme to Tejunic.
-->
Tejunic, here call it names.
-->
The output between the theme and Tejunic names.
-->
Unique players.
-->
And unites Tejunic with the THF output of the Unique.
-->
In other words, those who filtered the repeated ones will not add it again.
-->
Perfect.
-->
I hope then some are configuring.
-->
Perfect, ready.
-->
Then we are going to initially configure the HDFS input.
-->
This one that is here.
-->
As the HF input, what I want to read is the file called team NDA.
-->
Then please do the following.
-->
We are going to copy a metadata so that the scheme is easier for us.
-->
If there are some that have already downloaded it, those who do not have it downloaded, go to their Hadood cluster and please download the file to get the metadata.
-->
Download.
-->
And then it will create a metadata of the Limiter file that points towards that file that we download now.
-->
I'm going to call it here team NDA.
-->
And I think it's separated by a comma, I don't know, I have to check to see.
-->
I'm going to look for it.
-->
It must be in downloads.
-->
There it is, team NDA.
-->
Separated by comma, indeed.
-->
Separated by comma.
-->
And refresh.
-->
Team NDA.
-->
Team NDA.
-->
Yes, indeed, you have to mark it in header.
-->
And there it is.
-->
Perfect, ready.
-->
So then give it here if you tell me who is talking to me.
-->
Alejandro, I'm going to go to you, Alejandro.
-->
I don't know if you have it there.
-->
Wait a moment please.
-->
Yes, download it please.
-->
Click there.
-->
It just has the name there.
-->
Give it download.
-->
There it is on the left.
-->
There is the download.
-->
And now if you point it towards it.
-->
Brenda, I'm with you.
-->
Brenda, yes.
-->
Let's see.
-->
Give it there in refresh for a moment.
-->
No, it's weird.
-->
Give it back for a moment.
-->
Give it back.
-->
Give it...
-->
Give it next again to see.
-->
Wait a moment to see there.
-->
This is really weird here.
-->
What happened?
-->
Give it cancel for a moment.
-->
I'll open it again.
-->
To see if it was that.
-->
Give it to team NDA.
-->
Browse there.
-->
Ready.
-->
Give it to team NDA again.
-->
One moment, one moment.
-->
It's that the file is empty.
-->
It's in zero.
-->
Because it's in zero.
-->
Go to the folder for a moment.
-->
You were the one who downloaded the shared folder, right?
-->
Go to the shared folder for a moment to see.
-->
If you want.
-->
Go to the shared folder.
-->
If it was that I uploaded it wrong.
-->
Give it download up there.
-->
And I check to see if it has data in the shared folder.
-->
Because...
-->
Yes, it has data.
-->
This is to see how it looks now.
-->
Ready.
-->
Open it to see.
-->
Go to downloads.
-->
Now if it has data.
-->
You have to upload it again.
-->
But let's go here in the team.
-->
Let's go again.
-->
Yes, because it is in zero.
-->
You have to realize that it is in zero.
-->
Now upload it.
-->
Then take team again.
-->
Team NDA.
-->
Now yes.
-->
Next.
-->
Separated by comma and they have header.
-->
Give it to Sexhear.
-->
If you want to put...
-->
Perfect.
-->
Give it to refresh.
-->
There is a small problem.
-->
What happens is that I add one more when I click it.
-->
Give it.
-->
Put one there.
-->
One.
-->
Put the number one there.
-->
Give it again to refresh.
-->
Now yes.
-->
Give it to Next.
-->
Please.
-->
We give it to Next here.
-->
And right here.
-->
We take the button that allows us to export.
-->
Which is the penultimate.
-->
And I'm going to put it as usual.
-->
Temporal.
-->
And I'm going to call it Team NDA.
-->
XML.
-->
NDA.
-->
XML.
-->
Finish.
-->
Release.
-->
Release.
-->
Yes, because they have header that is empty.
-->
Well.
-->
Now we are going to confirm the HDFS input.
-->
First.
-->
Import the Team NDA scheme.
-->
It already has it.
-->
OK.
-->
Second.
-->
Configure the existing connection.
-->
Third.
-->
Refer to the Team NDA file.
-->
Where it says file name.
-->
Search it please.
-->
Team NDA.
-->
There it is.
-->
And the field separator is comma.
-->
And the header is one.
-->
Remember.
-->
I repeat again.
-->
Import the Team NDA scheme.
-->
Second.
-->
Refer to the file.
-->
AXA Team NDA.
-->
Third.
-->
Change field separator from point comma to comma.
-->
And tell it that it has a header.
-->
Act then.
-->
Ready.
-->
As we are going to configure the theme.
-->
And in the theme what we are going to do is simply make a column filter.
-->
To the theme you are going to take the following fields.
-->
One moment.
-->
I'm going to take.
-->
Player ID.
-->
I release it on this side.
-->
And when I release the first one I can expand here.
-->
It takes player.
-->
And it takes team.
-->
Three fields that are practically united.
-->
Now in the case of Haydn does not handle primary keys.
-->
Conventionally as we handle them in.
-->
Transitional database.
-->
But here when you are going to export to a database.
-->
For example.
-->
If you put here key.
-->
And you are going to send it to a database connection.
-->
It is what is going to try is that it is going to create this as a primary key.
-->
But Haydn does not handle primary keys.
-->
We call here OK.
-->
Yes.
-->
And we go here where it says Tejunik row.
-->
And we are going to configure the Tejunik row.
-->
Here it configures those elements in which it wants to eliminate the duplicates.
-->
It turns out that the player ID is going to appear n times.
-->
So there.
-->
It is mentioned in the section that says key attribute.
-->
The other section does not apply because player ID is whole.
-->
And the sensitive case only applies if it is textual.
-->
So what would I do there?
-->
I would only let the player ID pass once.
-->
If it is repeated, it does not let them pass.
-->
Ready.
-->
There is nothing else here.
-->
I'm going to the HDFS output.
-->
I say existing connection.
-->
Here I make sure that the field separator is point and comma.
-->
Because the one on the table is point and comma.
-->
I'm going to check the one on the table that is point and comma.
-->
Yes. Point and comma.
-->
Of the table.
-->
We are not going to change there.
-->
Here what I am going to reference is the name that I want to give practically.
-->
So I choose the connection first.
-->
And then I'm going to choose.
-->
Any name to then change it.
-->
So.
-->
And here I'm going to change it.
-->
And I'm going to call players CSV.
-->
players.csv
-->
Yes, just in case we make a mistake.
-->
Yes, it is true.
-->
Let's go here.
-->
Just in case we make a mistake.
-->
Please copy the name of the file.
-->
The output to paste it here in the Teja and Loat.
-->
I'm going to copy it here.
-->
I'm going to include header.
-->
Thank you.
-->
Say include header.
-->
Perfect.
-->
So include header.
-->
We are missing Teja and Loat.
-->
Teja and Loat says use existing connection.
-->
I choose the connection.
-->
The path will paste the output output of the HDFS.
-->
And the name of the table must coincide with the name that you create in the Teja and Loat.
-->
I think I call it players.
-->
Yes, players.
-->
And I paste it here.
-->
Please then join the two sub jobs with a sub job.
-->
Ok.
-->
And we have to test if the first point does not work then.
-->
Perfect, ready.
-->
Well, let's try it then.
-->
Make sure that the one who reads it does the job.
-->
I'm going to run to say there is no error.
-->
The output of the HDFS.
-->
Well, now you have to check that this table has data.
-->
You have two ways.
-->
Or you do a job.
-->
That lists the table.
-->
You know how it would be.
-->
It would be to update the connection.
-->
Extract and do a job.
-->
This is one way.
-->
The other way is to do the SELE directly in the console of the container.
-->
So, look at the following.
-->
If it worked well, take a screen.
-->
Place.
-->
First point.
-->
This screen and the screen that solves that has data.
-->
Or you do it with a SQL query.
-->
Or you do it with a job.
-->
If it is with a SQL query, you go here to the container.
-->
And you are going to put SELECT.
-->
Asterix.
-->
From players.
-->
And you take the screen.
-->
That's it.
-->
Ready.
-->
Here.
-->
Ok.
-->
Perfect then.
-->
I'm going with Beatriz.
-->
And the others already know what to do.
-->
I'm going with you Beatriz.
-->
One moment.
-->
While the others are placing.
-->
Let's see.
-->
Ok.
-->
Give it.
-->
Cancel the same moment.
-->
Go down please.
-->
Go down a little.
-->
Go down the job.
-->
Ok.
-->
In that text there is CLOS.
-->
Give it a moment.
-->
No.
-->
It is in the text HILOA.
-->
Give it a click in the text HILOA please.
-->
There he forgot to choose the existing connection.
-->
There is the empty combo.
-->
It says COMPON ELEASE.
-->
It did not choose the connection.
-->
Ok.
-->
Perfect.
-->
Now try to run to see.
-->
Ok.
-->
It is starting to run.
-->
Remember.
-->
It would be two screens.
-->
The job screen.
-->
And the screen that shows you the data.
-->
Of the table you created.
-->
The table or the consult.
-->
Ready.
-->
Ok.
-->
I'm going.
-->
Give me a moment.
-->
To finish with Beatriz.
-->
Beatriz.
-->
You already.
-->
Give him a cell from players there.
-->
Please you.
-->
Who spoke to me?
-->
Alejandro.
-->
Yes.
-->
I'm going with you for example.
-->
Select Asterix from players.
-->
Point comma.
-->
Ok.
-->
There it would have.
-->
I'm going with Alejandro then.
-->
Take the two screens please.
-->
In the file.
-->
Ok.
-->
Go then again to your job.
-->
Show me your output.
-->
Perfect.
-->
There you have it.
-->
Ok.
-->
In case of doubt.
-->
Go.
-->
Edit the SDC scheme.
-->
Put a moment there.
-->
Perfect.
-->
Ok.
-->
In case of doubt.
-->
Go to your.
-->
To your.
-->
Cluster.
-->
Leave your moment.
-->
And it should have the player file there.
-->
Go to browser.
-->
Give it F5 there.
-->
It's not there.
-->
I don't think so.
-->
Let's go again.
-->
Let's go again.
-->
Let's go again.
-->
To the.
-->
To the.
-->
To the job.
-->
That's weird.
-->
Go to the high loader for a moment.
-->
This is weird because the file doesn't have it.
-->
Go again to see what happens.
-->
Show me again.
-->
The cluster.
-->
In the browser.
-->
To see the moment.
-->
I didn't see the player file.
-->
It was there.
-->
You don't know it's not there.
-->
Ok.
-->
Go again to the job.
-->
Ok.
-->
That's it.
-->
Show me again.
-->
Show me again the job.
-->
Ok.
-->
Now check the table for a moment please.
-->
But you're going to have the nulls initially.
-->
And then you're going to have.
-->
Ok.
-->
Click there.
-->
Go.
-->
Execute.
-->
Yes.
-->
That.
-->
The null is still sending it.
-->
How so?
-->
Show me.
-->
Now yes.
-->
Show me the file.
-->
How did it stay in Hadoop for a moment.
-->
In the browser.
-->
It's at zero.
-->
Do you see it?
-->
It's at zero.
-->
It's at zero.
-->
Do you see it?
-->
Let's go again to the job.
-->
Go to the Unicron for a moment.
-->
Ah, I know.
-->
Go to the Tmac for a moment.
-->
There's the problem.
-->
Ah, no, no.
-->
There's the problem.
-->
There it is.
-->
Player ID.
-->
Ok.
-->
No, it's fine.
-->
Wait a moment.
-->
No, no, no.
-->
No, no.
-->
There's a problem.
-->
It seems like.
-->
Yes, do it again.
-->
No, it's fine.
-->
It's fine.
-->
Ok.
-->
Give it ok there.
-->
Go again to the Unicron.
-->
Go to edit scheme for a moment.
-->
Edit scheme.
-->
There it is.
-->
Give it ok.
-->
And go to the THF output at the end.
-->
Oh, it's interesting.
-->
Let me tell you that I didn't see the error.
-->
Let me tell you.
-->
I'll see.
-->
Well, let's do the following.
-->
In this case, I'm going to test this theme here.
-->
And I'm going to put a telor row here.
-->
And tell me.
-->
Can't you see it or don't you have it?
-->
Ah, then.
-->
Connect and give it.
-->
Use AXA.
-->
In.
-->
In Dock.
-->
It seems to be connected to the default database.
-->
Use AXA.com.
-->
And then it gives.
-->
Be a player to see.
-->
Let's execute it for a moment, Alejandro.
-->
If you don't wait a moment.
-->
Let's execute it for now.
-->
Let's execute it.
-->
Let's see.
-->
Why are you saving that data?
-->
Well, the data goes there.
-->
I mean, there are the data.
-->
I mean, no.
-->
If it happens.
-->
I'm going back here.
-->
What was it?
-->
It was here.
-->
Here.
-->
F5.
-->
But he deleted it, right?
-->
Tabla la crea.
-->
But this man.
-->
It seems.
-->
Players.
-->
F5.
-->
Alright.
-->
Well, let's do the engineering.
-->
Let's see.
-->
Let's remove this component from here.
-->
I'm going to remove this other one.
-->
And I'm going to check that at this point there is data.
-->
I mean, because suddenly the line can arrive.
-->
But with the empty data that is null.
-->
Data is coming there.
-->
So.
-->
I'm going to create another HDFSUP here.
-->
In case suddenly there was some configuration that we did not see.
-->
Let's connect it here.
-->
Ready.
-->
Use existing connection.
-->
I already copied the name.
-->
And this would be ready.
-->
Axa, players, text file.
-->
Edit scheme here.
-->
This scheme here.
-->
Point and comma.
-->
Because this man was with a point.
-->
But you say comma.
-->
It fits.
-->
As you have comma here.
-->
And the table is created.
-->
You have to put here.
-->
Comma then.
-->
Include header.
-->
Ready.
-->
Let's see now.
-->
The other would be to delete the table.
-->
To be able to change the separator.
-->
But since the table is created.
-->
The separator by default was comma.
-->
Let's see here.
-->
It would be here.
-->
Indeed, Alejandro.
-->
It was the small comma that we had there.
-->
Ready.
-->
Let's take the screens.
-->
I think it was.
-->
Thank you.
-->
Sometimes you do not realize those errors.
-->
Sometimes you do not realize.
-->
I made a mistake in mine.
-->
Who told me?
-->
Beatriz.
-->
Was it you who did not ...
-->
It was perfect.
-->
Then the first point is ready.
-->
Who was that?
-->
Ah, Fer.
-->
This is the one I said.
-->
I thought it was.
-->
Ready, let's see.
-->
Ah, no, it is not connected.
-->
Let's see.
-->
Go to your container, please.
-->
If you are going to talk to the stock docker.
-->
There is the one.
-->
The little whale.
-->
Give hype there.
-->
Hi.
-->
And you have to reconnect.
-->
Let me see if you have something up there.
-->
I do not know if they have the connection up there.
-->
That one that connects with slash.
-->
Or you copy it.
-->
In the folder to share.
-->
Share.
-->
Go there, ready.
-->
Expression.
-->
Enter expressions.
-->
Copy the connection you have there.
-->
That one.
-->
Paste it, please, there.
-->
No, with right click.
-->
No, right click.
-->
You have to right click.
-->
Ready.
-->
Paste there.
-->
Enter.
-->
Ready.
-->
Enter.
-->
Enter.
-->
Give it.
-->
Use.
-->
Use space AXA that would be the database.
-->
Space AXA dot com.
-->
And go straight.
-->
So give it select asterisk from players to see.
-->
It separates.
-->
And sometimes.
-->
From.
-->
Players.
-->
Dot com.
-->
There it is.
-->
What it does is that you were.
-->
Is that I think you were connected to the default database.
-->
So it was not there.
-->
So.
-->
Screen.
-->
Two screens.
-->
The first screen is that of the job.
-->
And the second screen is that it checks that the data is there.
-->
We continue.
-->
Ok.
-->
Well.
-->
Let's see.
-->
Who spoke to me?
-->
Brenda.
-->
Or Beatriz.
-->
Brenda.
-->
Let's see how Brenda moves.
-->
Let me.
-->
And we solve the problem.
-->
Please go to.
-->
To the job.
-->
Now.
-->
As it happened to us with Alejandro.
-->
Go to the Teja and Create Table moment.
-->
To check Teja and Create Table.
-->
Look to see the separator that has in the Teja and Create Table.
-->
Double click.
-->
And there it is separated by comma.
-->
Effectively what it has.
-->
And then as it separated by comma.
-->
Go to the HDFS output.
-->
And I would have to separate that file by comma.
-->
So that's where it says field separator.
-->
I would have to see comma because it had to be the same as the table.
-->
Please execute it again.
-->
Perfect.
-->
Go now to the Docker desktop.
-->
Give it.
-->
Up.
-->
Enter.
-->
There are going to appear the nulls but in the end they should appear with data.
-->
There are the nulls.
-->
Done.
-->
No, in the Word document.
-->
Place a first point and place the two images.
-->
So that in the end they send it to me by email that file.
-->
Well then we go with the next point.
-->
Which is to create the dimension time.
-->
So for this.
-->
I leave this open to copy the pre-yog and the post-yog.
-->
So I'm going back to Casio Studio and I'm going to create a new job here.
-->
I'm going to call it Job.
-->
Perfect.
-->
Alejandro.
-->
Done.
-->
Beatriz.
-->
I'm going to my screen now.
-->
Well then I'm going from here to Casio Studio and I'm going to create a new job.
-->
Job.
-->
Underscore.
-->
Dimension.
-->
Underscore.
-->
Time.
-->
Done.
-->
I'm going to copy the pre-yog.
-->
With all the elements it has.
-->
And I'm going to copy the post-yog.
-->
And finally I copy the input.
-->
Let the input be the same.
-->
Input.
-->
Copy here.
-->
And here.
-->
Perfect.
-->
Done.
-->
Yes, I need the input.
-->
Wait a moment and I'll review the file.
-->
This team.
-->
Let's see how it is.
-->
Freelance.
-->
Clients.
-->
Team, team, team.
-->
Moment.
-->
Wait a moment please.
-->
Yes, I need the input.
-->
I'm going to help here.
-->
A T-Map.
-->
A T-Unit.
-->
Ok.
-->
A T-Unit.
-->
And I finish with another.
-->
Lie.
-->
Better here.
-->
A T-Unit back here.
-->
More efficient here.
-->
A T-Map.
-->
Nothing else.
-->
And a T-HDF-C output.
-->
And a T-High-Low.
-->
It would be in those components.
-->
Done.
-->
Let's proceed to connect the elements.
-->
The T-HDF-Input we connect with the.
-->
With the T-Unit-Rob.
-->
The T-Unit-Rob.
-->
We connect it with the Units with the T-Map.
-->
And we connect the T-Map with the HDF-C output that is here.
-->
I'm going to call it here.
-->
Output.
-->
Formatted.
-->
Time.
-->
Perfect.
-->
Done.
-->
Remember that at the data warehouse level.
-->
Normally one makes a time dimension.
-->
And that time dimension takes out the different times from the data.
-->
So it generates a.
-->
A key that we call a synthetic.
-->
Some of it depends on the author.
-->
There are authors that are different.
-->
It connects the two sub-jobs.
-->
Or a sub-jobs.
-->
As we did last time.
-->
A sub-jobs.
-->
Let's configure the table first.
-->
Remember that the character.
-->
Of the table separation with the character of the output.
-->
By default here I have it with point and comma.
-->
But there are some that have it with comma.
-->
So everyone decides how to do it.
-->
The table is going to be called times.
-->
And in the blur scheme that it has.
-->
And we are going to add some columns.
-->
Some columns then.
-->
The columns are the following.
-->
Identifier.
-->
Of whole type.
-->
Identifier.
-->
Whole type.
-->
Month.
-->
Of whole type.
-->
Month.
-->
Whole type.
-->
Year.
-->
But it is going to have a problem with the year.
-->
So I put annuality.
-->
Whole type.
-->
And I put now.
-->
Month.
-->
Trimester.
-->
Whole type.
-->
And finally, name of the month.
-->
Name of the month.
-->
Stream type.
-->
There are five fields.
-->
Identifier.
-->
Month.
-->
Annuality.
-->
And trimester.
-->
All those four that I mentioned.
-->
Whole type.
-->
And the name of the month.
-->
Defines it.
-->
Stream type.
-->
It gives it here ok.
-->
If the impulse took it from the previous job.
-->
There is nothing more to do.
-->
And we are going to configure then the Tejunik row.
-->
And here what it is going to do is that.
-->
Wait.
-->
What impulse did I take?
-->
What a shame.
-->
I took the sales impulse.
-->
It is the one of players.
-->
I think I am the input.
-->
It is this input here.
-->
I copied.
-->
It was the different job.
-->
I paste here.
-->
I connect here.
-->
And it is going to eliminate the duplicates of dates.
-->
So.
-->
It stayed the same.
-->
What was it that I took the same?
-->
One moment.
-->
Ok.
-->
This has that it did not synchronize me.
-->
It left me the previous one.
-->
So I'm going to eliminate all these that are here.
-->
I'm going to eliminate them.
-->
Because I was wrong.
-->
So.
-->
Ready.
-->
And I pass all these here.
-->
Now yes.
-->
Ok.
-->
I was wrong.
-->
So here you have to filter.
-->
The dates.
-->
Act then.
-->
Well, let's hope there are some that are still.
-->
Configuring.
-->
Well, those who are already.
-->
There are some that are still configured.
-->
Please.
-->
Connect the two subjobs that I escaped now with 11 subjobs.
-->
Ok.
-->
There are some that are still configuring some things.
-->
With.
-->
Sorry.
-->
Connect the HDF input with the Teja and Lua with a subjoke.
-->
And the date.
-->
Yes sir.
-->
The date.
-->
Nothing more the date.
-->
I want to filter the dates that are duplicates.
-->
Yes.
-->
Only from there.
-->
Let's go then.
-->
Yes.
-->
No.
-->
One moment.
-->
Who brings me, right?
-->
Yes.
-->
Let's go then.
-->
We are in.
-->
Where is it?
-->
It is in the Tejunic Road.
-->
Pass all.
-->
In the Tejunic Road you have to pass all.
-->
Exactly.
-->
You must accept.
-->
Now.
-->
What is there.
-->
And now.
-->
Yes.
-->
Go to Tejunic and nothing more.
-->
Cool.
-->
Date.
-->
Perfect.
-->
Ready.
-->
Well, let's do the following.
-->
We leave it here.
-->
We continue now between one hour while we have lunch.
-->
And stand up so that some finish some things.
-->
So we return in one hour and we finish so that I can send that by mail.
-->
So good luck then.
-->
Yes.
-->
Good afternoon again.
-->
I hope you already have a good time in your lunch.
-->
If we are connected.
-->
Perfect.
-->
We are almost finishing then.
-->
All ready.
-->
We are going to continue.
-->
With the second point of the study case.
-->
Which comes now.
-->
We already filter.
-->
We remove the duplicates of the dates.
-->
And now each date must have a specific code.
-->
Since the dates are unique.
-->
There will be no dates repeated there by the Tejunic Road that is dedicated to that.
-->
Let's go to the theme at some point.
-->
The theme.
-->
We are going to configure the theme.
-->
And we are going to define a sequence that will place us each date.
-->
Which is unique with a unique identifier.
-->
For this part then.
-->
We are going to define a new variable here.
-->
Of full type.
-->
That we are going to call identifier.
-->
Identifier.
-->
And now I could then with sequences.
-->
Which is an option that has the talent.
-->
Which we already did before.
-->
Let's go then to the expression.
-->
And we look for where it says talent in numeric.
-->
There is an option that says sequence.
-->
Which is the last.
-->
And there it specifies the name of the sequence.
-->
Where it starts.
-->
And with what increment.
-->
It would start in one with an increment of one then.
-->
I'm not going to make any changes.
-->
Sequence is called.
-->
It is the last option.
-->
It gives you double click and it comes out automatically.
-->
Perfect.
-->
Perfect.
-->
Okay.
-->
I need now then format the dates.
-->
The dates that are here.
-->
Go to the entry elements.
-->
That here is the date.
-->
Here it specifies the format of the date.
-->
In which it is DDMMJJJ.
-->
But now it's my turn then.
-->
Take out the year.
-->
Take out the number of the month.
-->
Take out the quarter.
-->
Then we are going to do it the following way.
-->
We are going to create a new variable.
-->
Entire type.
-->
And we are going to call it then annuality.
-->
For the problem of the year that we have.
-->
Annuality.
-->
And we are going to take out the year from the date it has as an entry.
-->
For this then we are going to look for a function.
-->
That is called.
-->
Go to Talent Date.
-->
The functions that are grouped in the Talent Date category.
-->
There is already an option that is called.
-->
Get Part of Date.
-->
It's called like that.
-->
The function is called Get Part of Date.
-->
Double click on that function then.
-->
And the function is telling you what modifiers you should send.
-->
Depending on what you want to take out.
-->
As I want to take out the year.
-->
I'm going to send you year in capital.
-->
I can copy it from here.
-->
The same.
-->
Of the help that is here.
-->
And where it says, Date of Weeb.
-->
I replace it by year.
-->
What do I want to get?
-->
Then.
-->
I'm going to erase the example that is there.
-->
Which begins from the T.
-->
To the parenthesis.
-->
To the first parenthesis that closes.
-->
And I would stay comma.
-->
Parenthesis that closes nothing else.
-->
That's how it would stay.
-->
Comma and parenthesis that closes.
-->
y busco en dentro de la variable de entrada busco la que dice road dos date le doy doble
-->
click y me aparece justamente donde está el cursor y ahà sacarÃa el año perfecto le
-->
okey voy a sacar ahora el número del mes entonces para el número de mes voy a crear
-->
nueva variable voy a usar la misma función talent day get part of day es más creo que
-->
me sale más rápido que copie la expresión de la variable anterior y la pegue acá abajo sale
-->
más rápido y lo único que voy a cambiar es donde dice gear voy a colocar mom en inglés y en
-->
la expresión de la del año y pegarlo acá en la nueva variable que volverla configurar
-->
nuevamente a esto le damos okey vamos a cambiar el tipo sà dÃgame el tipo de datos serÃa entero
-->
primero porque la tal en devuelve un entero y el nombre serÃa mes me piden ahora el cuartel
-->
y para sacar el cuartel voy a hacer una función ya bastante antigua entonces voy a crear una nueva
-->
variable y voy a copiar la misma expresión del mes la copio y la pego en la expresión de acá
-->
vamos a ver dele hay más allá arriba más un momento un momento a ver qué pasa
-->
vuelvo a abrir nuevamente el tema no sé qué pasaba listo entonces para esto hacemos lo
-->
siguiente la función cuál serÃa vamos a restarle uno a la fech al mes que estoy
-->
sacando menos uno o copie brenda en su caso copie la expresión del mes en el talent day
-->
copia el anterior listo perfecto restele uno
-->
restele uno encierre entre paréntesis divida entre cuatro y sume le uno y esto saca esto
-->
me da el trimestre por ejemplo si tengo 11 entre cuatro entre cuatro darÃa dos
-->
sà no acuerdo si la función es entre tres o entre cuatro por un momento en la
-->
busco la función de se calcula la función entonces no sé si la función es para ver
-->
11 entre cuatro más uno no darÃa de quedar cuánto tenés es vamos a ver en 53 que pena
-->
entre tres no entre cuatro entre tres más un si entre tres más un
-->
esta función no sé si la estoy en el momento de buscarla apuntar rápidamente aquà a la
-->
ya dame la función a partir del mes que me del trimestre momento que no sé si me está dando
-->
es trimestre
-->
cuarto trimestre si señor esto sà está bien y listo sà entre tres entonces listo si perfecto
-->
ok y faltarÃa sacar aquà serÃa entero y serÃa aquà trimestre
-->
y creo que nada me falta el nombre del mes agrega aquà una nueva función pero aquà sà nos
-->
toca hacer una diferente porque la función está que par no tiene para obtener el nombre
-->
del mes entonces entre a la expresión por favor y va a buscar dentro de tal en date una
-->
función que se llama forma de y esta forma de este forma de forma de forma de entonces
-->
aquà en forma de donde está el formato que está ye ye ye vas a mandar como formato y
-->
4 m en mayúsculas que indique que quiero el nombre del mes m m m m 4 m en mayúsculas
-->
y en la opción que dice my day mandale como parámetro la variable que se llama road day
-->
la borra si tú haces el cursor en donde quieres le das doble clic a la ir creo que ya tenemos
-->
este punto aquà casi listo vamos a la que ok y antes de enviarlo de salida a perdón el
-->
nombre va a ser aquà nombre mes nombre mes se llama la variable y para ver el orden voy a
-->
ver cómo creer la tabla porque tengo que dar la misma orden la tabla está creada primero
-->
identificador número de mes anualidad trimestre y nomes el mismo orden entonces serÃa de esta
-->
manera mandarÃa primero va identificador la segunda columna va mes tercera anualidad
-->
cuarta trimestre y quinta nombre del mes y de nuevamente el orden identificador
-->
mes anualidad trimestre y nombre del mes porque lo coloque asà porque como voy a
-->
mandarla para que cargue el orden de los columnas tiene que ser igual a como definir
-->
la tabla y mande otro elemento aquà que se me escapó en la tabla que lo agrego
-->
ahora es sumamente importante que es la fecha para poder hacer un join ahà con esa fecha
-->
que se me olvidó de regar la tabla serÃa entonces identificador mes anualidad trimestre
-->
nombre el mes y date date como último campo entonces perfecto listo vamos a darle aquà ok ok
-->
agregue por favor en el texay create table agregue una nueva columna que se me escapó que
-->
era en el esquema agregue una nueva columna que se llame fecha y es de tipo date el tipo
-->
es tipo date ok
-->
bueno configuremos entonces fecha tipo de ahà ya nos falta el
-->
thdfs output entonces configurémoslo usar conexión existente
-->
listo entonces vamos al nombre archivo busque un nombre archivo que ya esté después lo
-->
llamamos en aksa el nombre archivo lo va a llamar tiempos.sv
-->
listo perfecto en el hay create table texay create table agregue un nuevo elemento que
-->
serÃa llámelo fecha de tipo date fecha de tipo date tipo date nada más fue ahà se
-->
estamos ahora en el output listo vamos en el output estamos confiando el output
-->
y el output serÃa entonces el nombre lo mande hacia aksa tiempos.sv le doy que sobrescriba
-->
y confirme que su separador de fila coincida con el separador que tiene en el create table
-->
yo por defecto tengo punto y coma y acá lo dejé en punto y coma y le dice que incluye
-->
perfecto entonces copie por favor la el nombre archivo del output para colocarlo
-->
con el texay load copiar
-->
vete a tu texay load en el file bueno usar conexión existente recuerde colocar aquÃ
-->
el nombre la conexión en el path y el nombre a tala se llama tiempos
-->
perdón se ve que venga vamos no vamos a configurar ya el texay load pegue por favor el
-->
path de la salida del output que serÃa aksa tiempos.sv y el nombre a tablas se llama tiempos
-->
bueno aquà creo que no hay que hacer más nada
-->
voy a probar ejecutar para ver si de pronto no tengo ningún tipo de error
-->
bien al parecer no hubo ningún error voy a ejecutar acá select asterisco frontiempos a ver
-->
bueno ahà están con los nombres noviembre listo voy a ver quién tiene problemas lo que les
-->
salió bien por favor recuerde punto número 2 un print screen del job y un print screen de la
-->
salida alejandro voy a ver qué pasó contigo a ver vamos a ver cuando ese es un error de
-->
compilación entonces para lo más seguro vete al si está en el texay load parece que hay una
-->
comilla doble que parece que leà para ver mira al inicio colocaste dos comillas dobles
-->
quÃtale una por favor ok eso creo que es el error ejecuta para ver
-->
perfecto dale un select asterisco por favor a la a la tabla tiempo y deberÃas tener los
-->
datos allá parecer fernanda creo que tiene un error voy a ver para ver bueno vamos a ver
-->
ese error de compilación vayase ok vayase al co para que lo marque le marque exactamente
-->
donde sacó que problema que tiene con su intercom no sé no sé por qué bueno vayase
-->
el punto rojo que tenemos ahà que le marca un error de compilación dice
-->
si quiere dele click acá en el en el rojito que tiene la izquierda si le coloca el mouse ahÃ
-->
le da click ahà de ese que tiene no puede convertir entero a string a seguramente el valor de la
-->
variable la colocó string vayase al tema y la la trimestre es entera pero espera un momento
-->
que aquà hay un problema cuando lo hago de aquà en entero y ahora enter de ley por favor ahora el
-->
trimestre de la salida o sea del lado derecho tiene que convertirlo también a entero en la
-->
parte de abajo están los mismos nombres de arriba que lo cambia ahà dice string y cambia
-->
intero dele para ver ejecutar
-->
por el creo que pronto colocate es colocate sobre escribir al té al loco de pronto que
-->
saca que se ha ido a loco en la acción dale sobre escribir porque parece que como que está
-->
intentando en el caso tuyo fernando a darse la catérica no era como con jaca con jorge en
-->
la que está rico tu fernanda ahora en la en la por favor en ahà en la acción jorge
-->
darle o sobre escribir ejecuta nuevamente para ver dale se la que está rico por favor fron
-->
tiempos escribe que él se colocan aparte abajo hay un pequeño error ahà vamos a ver
-->
dale que ir un momento ahà no el rollo de ejecución de beta el designer momento
-->
jorge qué pena contigo fernando te lo tiene bien voy con jorge entonces voy contigo ahora a ver
-->
cerca de y vete un momentico al tejai crédito por un momento
-->
le doy click ahà un momento a tu editar esquema
-->
ahà no sé por qué hubo un problema ahà ese tipo está bien pero el tipo de antes o sea el
-->
tipo que está ahà tenÃa que estar de ahà no sé por qué creo que lo cambiaste fue directamente
-->
en el debe tai si lo cambia el debe tai no te cambia el tipo de llave entonces ahora lo que
-->
no sé si creó la tabla entonces vuelve a ejecutar ahora se vuelve a ejecutar para ver
-->
no creo que yo me lo espero un momento y me conecto porque
-->
pero deja que termine un momento haz lo siguiente
-->
espérame y mira para ver si puede ver a mà me conecto interactivo para ver una configuración mejor
-->
ah ok aquà está un problema y creo que el inconveniente que tiene es que este campo
-->
no deja que sea no que se llame de y porque parece que es para la reservada
-->
un momento no no creo que vamos a ver para ver si es eso
-->
si deis lo tenÃa como para la reservada entonces hay y entonces ahà tenÃa el problema
-->
bueno ya está copiando vamos a ver
-->
tiempo serÃa tiempo tiempos creo que se llamaba tiempo perfecto listo copia si
-->
ahà tienes un problema con la última null de a tu momento yo no soy inconveniente ahÃ
-->
vete por favor ese no lo necesitamos ahora vete a tu a tu tema un momento a tu tema
-->
listo
-->
ahà está bien ok dale ok un momento ahà vete a tu hdfs output un momento
-->
y dale editar ahà ahà está bien también
-->
permÃtame yo tomarlo con todo el momento a ver y este es el hdfs que está acá
-->
si está bien vamos a hacer lo siguiente voy a borrar la tabla un momento drop table para
-->
que lo acabe nuevamente a ver si el problema sigue siendo ese perfecto listo vamos a ver
-->
entonces vamos a ejecutar nuevamente ese job tuyo
-->
está bien todo no lo veo nada extraño vamos a ver si
-->
la volve a borrar nuevamente pensé que estaba el cdk ahà que raro
-->
ah no no alcanzamos a ver verdad que sà vamos a ver si de pronto espera que termine
-->
está en nulo a los demás no salió nulo verdad yo a mà no no recuerdo que confirmó para ver yo no
-->
ah no pero es que yo también a un momento ah espera un momento vamos a correr a ver a ver
-->
porque está nulo es que todos tenemos nulo vamos a correr un momento que sin eso es vital
-->
para hacer la tabla de hechos
-->
no no creo que sea por el nombre porque es por posición del de los datos que no sé
-->
yo no no no no revisé si yo tenÃa también salió nulo entonces vamos a ver a a ti se salió
-->
ah ok bueno ahora vamos a buscar el caso tuyo para ver porque a ti sà y a nosotros no
-->
no sé te haces confirmar porque cerré aquà el doc
-->
y este docker tiene un pequeño problema en esta versión de windows que cuando
-->
cierro la ventana de forma bruta no me la abre más ah pero es un momento
-->
ah
-->
en el caso tuyo un momento a ver fernanda es porque a ti sà y al demás no que
-->
voy a entrar a tu ventana para ver fernanda no en realidad a ti te falta una campo fernanda
-->
que es la fecha el último campo 10 5 son 6 si vamos a vamos a comprobar tu yo que un momento
-->
y el te hay que ir aquà deberÃa tener si aquà falta un campo creerÃa que es
-->
momento a ver dele un descrae la tabla por favor y mire qué tipo de datos lo le apareció
-->
que tengo un problema acá dele des haga lo siguiente váyase al váyase si es tan amable
-->
al al contenedor y coloque lo siguiente des des en este caso serÃa tiempos.com no la fecha está bien este tipo date
-->
si creo que va a ser el problema es que yo creo que haga lo siguiente para ver cómo esto es
-->
del siguiente select current
-->
con the record date el asà para ver creo que esto con
-->
no recuerdo si es con el asà para ver current date
-->
para ver qué formatos tiene por defecto año mes dÃa entonces
-->
voy a ver este day y este day está en
-->
dÃa mes año por favor aquà en el data pattern este del del tema cambiela el orden aquà para
-->
que sea dd m m y y dd guión m m mayúscula dd ok
-->
vamos a dar aquà yes vuelvo a ejecutar yo harÃa un docker sec
-->
docker
-->
docker
-->
And now, in the last ones, if I add them in all of them.
-->
Oh, well, then, that error is already a moment.
-->
So there is a problem that you have to format that date.
-->
The problem that we had then was that the date,
-->
the format of the date that I am sending is not the same
-->
format that the problem that is causing already has.
-->
But for a moment there, I do something here, a moment,
-->
please, this is here in documents, just a moment, please.
-->
Hey, I connect here, I think I have it here, I have already seen it.
-->
Well, it is clear that the format of the date does not like the
-->
date, then I have to define a different format.
-->
Ready to copy.
-->
Wait a minute.
-->
Select a 30 of them.
-->
Okay, let's do something then. Let's turn that date into a stream, better. Wait a moment,
-->
I'm thinking right. What happens is that I don't know if this format, as it is entering me in this way,
-->
when I pass it through here I lose the format completely. I'm going to run it again to see what
-->
it is giving me, how it is saving those data. Can you tell us please that you sent it as
-->
data? Okay, and the last one because it didn't work, let's see what happened.
-->
Wait a moment, let's see, I'm going to see the original file to see what's going on.
-->
In what format is the original file? Wait a moment. Ah, the original,
-->
it's a bit messy, so there's no problem. Here we have a problem, that is, the problem is in format, so
-->
that's it, I understand. So, let's do the following, let's do the following then,
-->
let's go to the input, the input, let's edit the input-input scheme and we are going to change
-->
the format from here in, the format we are going to read it then here in yy-mmdd.
-->
Okay, let's see if I changed the unit, they all change immediately. Yes, I think so, yes, they all change.
-->
And in the theme, let's see. Yes, then let's do that. If you want, please give a delete
-->
to the data in the table, give it delete from, delete from times and run again to see.
-->
Ah, no, this command doesn't leave me with delete, it leaves me with, wait a moment, where is the command?
-->
But hey, if it executed it, take the screens there, this man to save it is not with delete.
-->
Well, I'm going to do a drop table better and run it again, give it a drop table, times,
-->
of speed because of the time that it is no longer consuming, delete it and run it again.
-->
And run everything again then, in this case I also run it again.
-->
And that conversion was strange then.
-->
Perfect, I think it's over, now if it's going to be a select, I'll select it.
-->
Well, in my case, I'm still null, I don't know why.
-->
DDD, okay, I don't know if I'm going to be able to relate to what it was. Well, I don't know, if the dates came out,
-->
we are going to be null, we are going to see what happened then.
-->
Ah, of course, edit the scheme, let's go then, the format here would be, we don't change it here, right?
-->
We are going to change it here then, let's see, here it would be,
-->
yymmdd, I don't know if, and this man, the dbtie has to be here date, I don't know why it appeared when it was here,
-->
here it has to be date, date, date and in this format, I don't know why I changed it there,
-->
delete the table again to see, because if not, it will not create with the changes.
-->
And ok, and we run again to see, you have to delete the table because if the change later does not affect, it does not take into account the change.
-->
Effectively, of course, now we are going to see, Beatriz, now we are going to see, let's go to the changes with you, Beatriz, to see them first.
-->
First change, let's go to the HDFS input, let's check the changes, go to HDFS input and verify that the format is yymmdd, in the scheme, perfect,
-->
yymmdd, enlarge a little more to confirm that it is well written, perfect, ok, then go to the test high create table,
-->
the one above, the create table, that one that is there, and give it to edit the scheme, please.
-->
Change there, that dbtie, I don't know why it changed, it must be date, not time-stan,
-->
that time-stans that it has next to the last row, it appears dbtie, in the column dbtie, time-stan, change it to date,
-->
I don't know why it changed to time-stan there, and now change the format to that one that is there, change the format there to yymmdd,
-->
yymmdd, ok, now go to the docker, and give it drop, space, table, times,
-->
I will execute the joe again if it is so kind,
-->
I don't think it needs anything there, I don't see anything there that it needs, execute to see,
-->
we are going to leave up to that second point, please save the document and send it to me
-->
that mail that I sent out there, please, I confirm it when I get there, I sent it by
-->
checking the mail, save the document up to point number 2, what they have already finished,
-->
go to the docker again, there it is, thank you very much, it was an interesting course,
-->
because you helped me a lot, that's why working with people who know how to do it is extremely
-->
interesting, well, there I sent you the mail so that you can send me that document that you worked on,
-->
they are two points that are going to have nothing else, that has to send me, yes, tell me,
-->
of course, of course, yes, of course, send it to me, but send it to me separately,
-->
please, so that later you confuse me, this is what we did now, perfect, ready,
-->
good question, good question, so look what it's going to be, it goes to where the talent is installed,
-->
which is the program file, one moment, let's see if I left it there, tost bd,
-->
and here there is a subfolder called workspace, enter workspace, and here is the project
-->
called local project, so imagine that you changed the project to project axa,
-->
in the end what you are going to do nothing else to have all the copy is to copy everything
-->
that is here within the name of your project and there it automatically recreates everything,
-->
look here, compress this folder, which is the important thing, and the content of that folder is the one that
-->
is going to paste in the project that you defined, by default local project is created, but everything
-->
that is inside is pasted into the local project and automatically you have all the exercises
-->
that we did there, yes, that too without needing the tools, exactly, well, here then,
-->
let me know when you send it, I will confirm it, please, the email,
-->
close it to have a copy of the exercises that we did, so if you want to send it,
-->
close your email and when you have the tool you can again recover those examples that we
-->
did in class, what you see send me by email is the doc document where it has point number 1
-->
and point number 2, which is the valuable one, the compressed one is to have a copy of the
-->
example that we did, call me in your house, here I got González Puentes and Pesquera
-->
RodrÃguez, I already got it ready, I sent it here through the chat, there in that email you can
-->
write to me for any questions, you are going to write to me for the good practices and any
-->
questions you have, please write to me there, here I got Sánchez, Sánchez Rosas Beatriz,
-->
yes, I also got it, Jimena arrived and I got Jiménez Ramos and Rosales Trejo, perfect,
-->
I think they all sent it, right? I think they all sent it, well, without a doubt, I did not
-->
make a mistake in the introduction when I said that it was a transfer of knowledge of this
-->
account that many of you helped me in some problems and I really loved having
-->
worked with you, I wish you the best, any questions write me, I am willing to
-->
if you have any questions about the topic of such or what things write to me and if I have the
-->
opportunity to continue the knowledge, I will gladly share it with you, see you then,
-->
see you then, it has been a pleasure, I wish you the best, happy rest of the day then,
-->
happy weekend and have a great time then, yes, tell me, Muniz has not arrived, Muniz,
-->
where did you send it? It has not arrived yet, did you just send it or did you send it? González Fuentes,
-->
Pesquera RodrÃguez, Sánchez Rosas, Sánchez Rosas sent it to me, ah, he sent me another question, Jiménez Ramos,
-->
no, it has not arrived, did you send it to John Jaime? Yes, let's see if I have it, I don't think I have it,
-->
in Spain I don't think I have it, no, send it back again to see, sometimes it takes a while, but
-->
share the screen, please, to see the name of the email, let's see if I'm going, yes, let's see,
-->
send it back, please, yes, ah, ready, yes, what a strange question,
-->
how strange, I don't know why it's strange, because they were ordered, I gave it F5 and he put it in the third,
-->
that is, he inserted it and it happened twice, not twice, well, yes, it has already been confirmed,
-->
well, I wish you the best, great group, have a great time,