Pull data from a table hosted in pltaform.sh
The information in this post is accurate as of the published date . Please make sure to check any linked documentation within the post for changes/updates.
I want to query a maria db table a hosted on Platorm.sh and bring the data to AWS . How to do this ?
-
If you are using our CLI, you can do
> platform sql -e <environment-name> -p <projectid>
which will drop you into a myql prompt.
If you just need a local dump of the entire database, you can run
> platform db:dump -e <environment-name> -p <projectid>
which will create a local
*.sql
file containing the contents, and from there should be able to import that into another system.0 -
Hi Paul Gilzow,
Thanks for your response.
But I do not want the entire database to be pulled. I want to programmatically pull data from a single table. For eg: ‘select * from table1 where column1 = ’Y' . I want to use this SQL in a ETL tool or a Python script . How to establish this ?
0 -
the
platform sql
command also allows you to pass a sql command to the db instance. From there, you could redirect the output to a file. Something like:> platform sql -e <environment-name> -p <projectid> ‘select * from table1 where column1="Y";’ --raw > somefile.txt
alternatively, you can create a tunnel to the database service.
0 -
@Paul Gilzow
Again ,my question is, How to embed this platform sql statement in a Python script or ETL tool (eg: Informatica/Alteryx) and schedule it to run automatically from a remote server?
0 -
in terms of a specific ETL tool, I can't answer since I haven't used those tools.
As for python, you could either have python call the platform cli tool directly (
subprocess.Popen
). Or you could create a tunnel from the remote server to the platform db service and then connect to the db instance from your python code via the tunnel.You mention a remote server, so you could have a cron job that runs the platform cli command on a schedule dumping to a file, and then your python script (or maybe the ETL tools) could read from the file.
0
Please sign in to leave a comment.
Comments
5 comments