From: | Venkata Balaji N <nag1010(at)gmail(dot)com> |
---|---|
To: | Wei Shan <weishan(dot)ang(at)gmail(dot)com> |
Cc: | Raghavendra <raghavendra(dot)rao(at)enterprisedb(dot)com>, pgsql-admin <pgsql-admin(at)postgresql(dot)org> |
Subject: | Re: Moving pg_xlog folder |
Date: | 2015-06-29 04:19:35 |
Message-ID: | CAEyp7J-V1iKOaw+pxp4wWurPJuQsW2=bBw+9-=6yeTxTdkra8A@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
>
> It seems that even when the port was configured in postgresql.conf to be
> 5433, it was still running in 5432. After I reconfigured the port to be
> 5432, the steps are working fine already.
>
Or, you can say "psql -p 5433" - that can help you connect successfully.
Good to know that everything is working fine. When you run "psql" without
any options, it would by default take the port as 5432 or whatever port you
have assigned to PGPORT environment variable. If you want to connect to
PostgreSQL with a different port and PGPORT is not configured, then you
need to exclusively mention the port with -p option.
Regards,
Venkata B N
Fujitsu Australia
>
>
> On 28 June 2015 at 14:43, Raghavendra <raghavendra(dot)rao(at)enterprisedb(dot)com>
> wrote:
>
>> On Sun, Jun 28, 2015 at 12:19 AM, Wei Shan <weishan(dot)ang(at)gmail(dot)com> wrote:
>>
>>> Hi all,
>>>
>>> OS: CentOS
>>> PG version: 9.3
>>>
>>> I'm testing the scenario of moving the xlog folder to a fast pair of
>>> disk. This was the steps I followed;
>>>
>>> 1. create new mountpoint with the correct permissions
>>> 2. Shutdown DB
>>> 3. Create softlink from new mountpoint to $PG_DATA/pg_xlog
>>> 4. Startup DB
>>>
>>>
>> If I understand here correctly, you want to move $PGDATA/pg_xlog to
>> "/fast/disk/pg_xlog" directory.
>>
>> If I want to implement the same on my local machine of moving of pg_xlog
>> to other directory (in your case other mount point). I will follow below
>> steps:
>>
>> 1) Stop the running cluster
>>
>> -bash-4.2$ pwd
>> /opt/PostgreSQL/9.4
>> /opt/PostgreSQL/9.4/bin/pg_ctl -D /opt/PostgreSQL/9.4/data/ stop -mf
>>
>> 2) Create a new directory (In your case new mount point with correct
>> permission)
>>
>> mkdir temp_pgxlog
>>
>> 3) Copy existing files and directory from current $PGDATA/pg_xlog to new
>> mount point directory.
>>
>> cp -rf /opt/PostgreSQL/9.4/data/pg_xlog/* /opt/PostgreSQL/9.4/
>> temp_pgxlog/
>>
>> 4) Create symbolic link to new directory by renaming existing directory.
>>
>> cd /opt/PostgreSQL/9.4/data/
>> mv pg_xlog/ pg_xlog_bkp
>> ln -s ~/temp_pgxlog/ pg_xlog
>>
>> 5) Start the cluster.
>>
>> /opt/PostgreSQL/9.4/bin/pg_ctl -D /opt/PostgreSQL/9.4/data/ start
>>
>>
>> However, after I start the DB, I hit the following error;
>>>
>>> *-bash-4.1$ psql*
>>>
>>> *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"?*
>>>
>>> This issue resolves by itself once I reboot the server after I move the
>>> XLOG. Any ideas what I did wrongly?
>>>
>>> Thanks!
>>> --
>>> Regards,
>>> Ang Wei Shan
>>>
>>
>> Reboot of the server not required, have you check below areas before
>> rebooting.
>>
>> 1) Is symbolic link of $PGDATA/pg_xlog pointing to new location ?
>>
>> 2) What information you see in the Database server logs ?
>>
>> ---
>> Regards,
>> Raghavendra
>> EnterpriseDB Corporation
>> Blog: http://raghavt.blogspot.com/
>>
>>
>
>
> --
> Regards,
> Ang Wei Shan
>
From | Date | Subject | |
---|---|---|---|
Next Message | Dave Johansen | 2015-06-29 04:45:01 | Re: database must be vacuumed with <N> transactions |
Previous Message | Wei Shan | 2015-06-28 10:12:18 | Re: Moving pg_xlog folder |