Re: Failed to run initdb - not resolved bug 5130

From: Robert Haas <robertmhaas(at)gmail(dot)com>
To: Tomas Studva <tstudva(at)gmail(dot)com>
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: Re: Failed to run initdb - not resolved bug 5130
Date: 2010-01-04 17:21:29
Message-ID: 603c8f071001040921p7b3b890aye2cbe07c8b559aab@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

2010/1/2 Tomas Studva <tstudva(at)gmail(dot)com>:
> Hi all gurus,
>
> I've encountered on my computer this bug. I will try to describe all what I
> have discovered till now.
>
> 1, this bug happened a few times before and wasn't resolved:
>   bug 5130 -
> http://www.mail-archive.com/pgsql-bugs(at)postgresql(dot)org/msg24810.html
>   bug 5222 - http://archives.postgresql.org/pgsql-bugs/2009-12/msg00004.php
>   bug 4691 - http://archives.postgresql.org/pgsql-bugs/2009-03/msg00028.php
>   ? - http://forums.enterprisedb.com/posts/list/1845.page
> I think the problem was always the same.
>
> My case is WIN XP SP 2 and also when applied SP3. My system has only one
> speciality as I know, that is I have windows installed on F partition.
>
> Problem description:
>
> I tried install with one click installer, it ends with error postgresql.conf
> not found. In log it is failing on initdb, which clears data dir after
> failure. Uninstaller is not working, do nothing when runned. When run
> installation again, then upgrade mode is run, because installation is found.
> This can be tricked by deleting installation from registry and deleting win
> service, then new install is possible.
>
> I was installing on path F:\Program Files\PostgreSQL\8.4 and also tried
> C:\Program Files\PostgreSQL\8.4, problem is probably not with install path.
>
> OK, what more I have tried. MS Visuall C++ redistr. I uninstalled, so the
> one with postgres comming were installed surely. I tried to enable debug
> output, but it gives no new information. I was reading initdb sources and
> problem is probably in this part, because I see message "selecting default
> max_connections " then one error message. It looks for me as snprintf is not
> executed or reached or etc.
>
> -----------------------------------------------------------------------------
>   printf(_("selecting default max_connections ... "));
>   fflush(stdout);
>
>   for (i = 0; i < connslen; i++)
>   {
>       test_conns = trial_conns[i];
>       test_buffs = MIN_BUFS_FOR_CONNS(test_conns);
>
>       snprintf(cmd, sizeof(cmd),
>                SYSTEMQUOTE "\"%s\" --boot -x0 %s "
>                "-c max_connections=%d "
>                "-c shared_buffers=%d "
>                "< \"%s\" > \"%s\" 2>&1" SYSTEMQUOTE,
>                backend_exec, boot_options,
>                test_conns, test_buffs,
>                DEVNULL, DEVNULL);

Are you able to compile from source? If so, can you hack this part to
print the value of cmd at this point, so you can see clearly exactly
what it's trying unsuccessfully to do?

...Robert

In response to

Responses

Browse pgsql-bugs by date

  From Date Subject
Next Message Robert Haas 2010-01-04 17:24:02 Re: BUG #5260: Postgre crashed unexpectedly
Previous Message Takesver Kumar Thakur 2010-01-04 13:44:21 BUG #5261: Invalide page header