Re: Server won't start with fallback setting by initdb. - Mailing list pgsql-hackers

From Robert Haas
Subject Re: Server won't start with fallback setting by initdb.
Date
Msg-id CA+TgmoZtx6xBCMhH8aYugN1gom2ya2e7BYoQRXuP8_DVyLMVWQ@mail.gmail.com
Whole thread Raw
In response to Server won't start with fallback setting by initdb.  (Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>)
Responses Re: Server won't start with fallback setting by initdb.  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Fri, Feb 9, 2018 at 3:08 AM, Kyotaro HORIGUCHI
<horiguchi.kyotaro@lab.ntt.co.jp> wrote:
> I'm not sure such a case happens in the real world nowadays,
> initdb uses the fallback value of max_connections=10. But it is
> out of favor of server since it is not larger than
> max_wal_senders(10).
>
>> postgres: max_wal_senders must be less than max_connections
>
> I think that we can safely increase the fallback value to 20 with
> which regtests are known not to fail. I believe that is
> preferable than explicitly reducing max_wal_senders in the
> generated config file. I confirmed that tegtest won't fail with
> the value. (Except with permanent failure of dynamic shared
> memory)

I propose an alternative fix: let's instead change the code like this:

        if (max_wal_senders > MaxConnections)
        {
                write_stderr("%s: max_wal_senders may not be more than
max_connections\n", progname);
                ExitPostmaster(1);
        }

That way, the behavior would match the documentation, which says:

"WAL sender processes count towards the total number of connections,
so the parameter cannot be set higher than max_connections."

Alternatively, we could change the documentation to match the code and
then do this.  But it's not good that the documentation and the code
don't agree on whether equal values are acceptable.

-- 
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company


pgsql-hackers by date:

Previous
From: Michael Banck
Date:
Subject: [PATCH] Verify Checksums during Basebackups
Next
From: Anastasia Lubennikova
Date:
Subject: Re: [HACKERS] Cast jsonb to numeric, int, float, bool