On 03/17/2017 09:49 AM, Adrian Klaver wrote:
> On 03/17/2017 06:42 AM, Steve Clark wrote:
>> Hi List,
>>
>> I am running postgresql 8.4.20 on CentOS 6. Things have been running fine
>> for a long time
>> then I rebooted. Postgres came up but when I tried to connect with psql on
>> the local machine
>> I got
>>
>> psql: could not connect to server: No such file or directory
>> Is the server running locally and accepting
>> connections on Unix domain socket
>> "/var/run/postgresql/.s.PGSQL.5432"?
>>
>> The socket is actually being created (and always has been) in /tmp
> So you built Postgres from source?
No - it is installed using yum.
It looks like the problem is someone loaded postgresql-9.5 and psql was using:
psql 14971 postgres mem REG 8,4 193296 950807
/usr/pgsql-9.5/lib/libpq.so.5.8
the libpq from 9.5.
By removing the ln and using
# rm /var/run/postgresql/.s.PGSQL.5432
rm: remove symbolic link `/var/run/postgresql/.s.PGSQL.5432'? y
$ psql
psql: could not connect to server: No such file or directory
Is the server running locally and accepting
connections on Unix domain socket "/var/run/postgresql/.s.PGSQL.5432"?
$ LD_LIBRARY_PATH=/usr/lib64 psql
psql (8.4.20)
Type "help" for help.
postgres=#
It works OK again.
So now I know what caused the problem.
Thanks,
Steve
>> I worked around the problem temporarily by
>> ln -s /tmp/.s.PGSQL.5432 /var/run/postgresql/.s.PGSQL.5432
>>
>> What controls where psql looks for the socket?
> https://www.postgresql.org/message-id/23876.1488949292%40sss.pgh.pa.us
>
> "With the default configure options you used, the postmaster would have
> put its Unix socket file into /tmp, not /var/run. I wonder whether your
> problem is that you're trying to connect to it with distro-supplied
> psql+libpq that expects to find the Unix socket in /var/run."
>
>
>
>> Thanks,
>> Steve
>>
>>
>>
>>
>
--
Stephen Clark
*NetWolves Managed Services, LLC.*
Director of Technology
Phone: 813-579-3200
Fax: 813-882-0209
Email: [email protected]
http://www.netwolves.com
