Re: Server Startup Error - Mailing list pgsql-performance

Note - try to cc: the mailing list, I don't always read this inbox

Gauri Kanekar wrote:
> On 2/26/07, Richard Huxton <dev@archonet.com> wrote:
>>
>> Gauri Kanekar wrote:
>> > Hi List,
>> >
>> > Machine was down due to some hardware problem.
>> >
>> > After then when i issue this command /usr/local/pgsql/bin/psql -l
>> > its giving me the following error
>> >
>> > psql: could not connect to server: No such file or directory
>> >        Is the server running locally and accepting
>> >        connections on Unix domain socket "/tmp/.s.PGSQL.5432"?
>> >
>> > Can anybody tell me what going wrong??
>>
>> Well, it's either looking in the wrong place or the server isn't
>> actually running.
>>
>> 1. Do your startup scripts start PG?
>
> Yes

OK - so we know it should have started, which means the logs should say
something about our problem.

> 2. Is there a server process? "ps auxw | grep postgres"
>
> This is the result given by the command
> root      8907  0.0  0.1  37496  2640 ?        Ss   03:47   0:00 sshd:
> postgres [priv]
> postgres  8910  0.0  0.0  37636  1684 ?        S    03:47   0:00 sshd:
> postgres@pts/1
> postgres  8911  0.0  0.1  10152  2564 pts/1    Ss+  03:47   0:00 -bash
> root      9470  0.0  0.1  37500  2644 ?        Ss   04:28   0:00 sshd:
> postgres [priv]
> postgres  9473  0.0  0.0  37640  1688 ?        S    04:28   0:00 sshd:
> postgres@pts/2
> postgres  9474  0.0  0.1  10104  2412 pts/2    Ss   04:28   0:00 -bash
> postgres  9724  0.0  0.0   3496   892 pts/2    R+   04:44   0:00 ps auxw
> postgres  9725  0.0  0.0   3868   784 pts/2    R+   04:44   0:00 grep
> postgres

Hmm - nothing there but "ssh" connections. So, it's not started, which
is why psql is complaining.

> 3. What do your logfiles say?
>
>
> HINT:  In a moment you should be able to reconnect to the database and
> repeat your command.
> LOG:  database system was interrupted at 2007-02-23 20:14:24 IST
> LOG:  could not open file "pg_xlog/00000001000000390000001A" (log file 57,
> segment 26): No such file or directory
> LOG:  invalid primary checkpoint record
> LOG:  could not open file "pg_xlog/000000010000003900000017" (log file 57,
> segment 23): No such file or directory
> LOG:  invalid secondary checkpoint record
> PANIC:  could not locate a valid checkpoint record
> LOG:  startup process (PID 9057) was terminated by signal 6
> LOG:  aborting startup due to startup process failure
> FATAL:  pre-existing shared memory block (key 5432001, ID 1900546) is still
> in use
> HINT:  If you're sure there are no old server processes still running,
> remove the shared memory block with the command "ipcclean", "ipcrm", or
> just
> delete the file "postmaster.pid".

OK - this last bit is the first thing to deal with. Find your
postmaster.pid file and delete it. Your postmaster.pid file should be in
your data directory - try "locate postmaster.pid" or "find /usr/local/
-name postmaster.pid".

Then restart postgresql (as root "/etc/init.d/postgresql start" or
similar) and check the logs again.

--
   Richard Huxton
   Archonet Ltd