Re: Executing stored procs

From: Daniele Varrazzo <daniele(dot)varrazzo(at)gmail(dot)com>
To: Anthony Waye <anthony(dot)waye(at)arq(dot)group>
Cc: "psycopg(at)postgresql(dot)org" <psycopg(at)postgresql(dot)org>
Subject: Re: Executing stored procs
Date: 2020-01-15 12:05:37
Message-ID: CA+mi_8ZhBBvo7pn4_EzJ2joBRXxYhygABz2RQ9AK7Vg0Wkd7eQ@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: psycopg

On Wed, Jan 15, 2020 at 7:37 AM Anthony Waye <anthony(dot)waye(at)arq(dot)group> wrote:

> Essentially results in a SQL query of: CALL
> "stg_customervip"."sp_stg_customer"(37::bigint, '2020-01-15
> 05:52:31'::timestamp)
>
> If I take that query and run it directly in redshift it runs successfully
> but via psycopg2 it returns:
>
>
>
> psycopg2.errors.FeatureNotSupported: TRUNCATE cannot be invoked from a
> procedure that is executing in an atomic context.
>
> HINT: Try calling the procedure as a top-level call i.e. not from within
> an explicit transaction block. Or, if this procedure (or one of its
> ancestors in the call chain) was created with SET config options, recreate
> the procedure without them.
>
> CONTEXT: SQL statement "TRUNCATE table stg_customervip.Customer"
>
>
>
> While that error sounds legitimate I think it might be a redherring
> because it does execute successfully if I do it manually against redshift.
>

Probably if you run it manually you do it outside a transaction. Psycopg
starts a transaction automatically (no, I don't think it's a good idea, but
it's part of the specs)

http://initd.org/psycopg/docs/usage.html#transactions-control

Try setting `redshift_conn.autocommit = True` after connection creation,
and do without the `commit()`s.

-- Daniele

>

In response to

Responses

Browse psycopg by date

  From Date Subject
Next Message Anthony Waye 2020-01-16 01:01:30 RE: Executing stored procs
Previous Message Anthony Waye 2020-01-15 06:36:26 Executing stored procs