Re: should we document an example to set multiple libraries in shared_preload_libraries? - Mailing list pgsql-hackers

From Bossart, Nathan
Subject Re: should we document an example to set multiple libraries in shared_preload_libraries?
Date
Msg-id DCA9F7EE-9F71-41F7-8537-74FA41A64957@amazon.com
Whole thread Raw
In response to Re: should we document an example to set multiple libraries in shared_preload_libraries?  (Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>)
Responses Re: should we document an example to set multiple libraries in shared_preload_libraries?  (Michael Paquier <michael@paquier.xyz>)
List pgsql-hackers
On 12/1/21, 5:59 AM, "Bharath Rupireddy" <bharath.rupireddyforpostgres@gmail.com> wrote:
> On Wed, Dec 1, 2021 at 6:45 PM Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> Considering the vanishingly small number of actual complaints we've
>> seen about this, that sounds ridiculously over-engineered.
>> A documentation example should be sufficient.
>
> Thanks. Here's the v1 patch adding examples in the documentation.

I think the problems you noted upthread are shared for all GUCs with
type GUC_LIST_QUOTE (e.g., search_path, temp_tablespaces).  Perhaps
the documentation for each of these GUCs should contain a short blurb
about how to properly SET a list of values.

Also upthread, I see that you gave the following example for an
incorrect way to set shared_preload_libraries:

        ALTER SYSTEM SET shared_preload_libraries =
        auth_delay,pg_stat_statements,sepgsql;  --> wrong

Why is this wrong?  It seems to work okay for me.

Nathan


pgsql-hackers by date:

Previous
From: "Bossart, Nathan"
Date:
Subject: Re: parse_subscription_options - suggested improvements
Next
From: Michael Paquier
Date:
Subject: Re: Skip vacuum log report code in lazy_scan_heap() if possible