From: | Vibhor Kumar <vibhor(dot)kumar(at)enterprisedb(dot)com> |
---|---|
To: | Eduardo Morras <nec556(at)retena(dot)com> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: UFS2 Snapshots and Postgres |
Date: | 2012-06-25 15:09:12 |
Message-ID: | 1398F3D4-87DA-4D42-B016-06804D4ED8F5@enterprisedb.com |
Views: | Whole Thread | Raw Message | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Jun 25, 2012, at 10:34 AM, Eduardo Morras wrote:
>>> Thanks in advance
>> Snapshots are "safe" (but will result in a roll-forward on restart) IF AND ONLY IF the log data and database table spaces are all on the same snapshotted volume.
>>
>> IF THEY ARE NOT then it will probably work 95% of the time, and the other 5% it will be unrecoverable. Be very, very careful -- the snapshot must in fact snapshot ALL of the involved database volumes (log data included!) at the same instant.
>
> Even if i do a pg_start_backup()? I thought it set db data/ files in a consistent state and puts in wal files the new transactions and apply them when call pg_stop_backup().
>
> I must do it other way then :(
>
> Thanks
If you are doing pg_start_backup(), taking snapshot and pg_stop_backup with archived WAL file backups. Then its safe.
Thanks & Regards,
Vibhor Kumar
EnterpriseDB Corporation
The Enterprise PostgreSQL Company
Blog: http://vibhork.blogspot.com
From | Date | Subject | |
---|---|---|---|
Next Message | Merlin Moncure | 2012-06-25 15:28:13 | Re: row_to_json question |
Previous Message | Tom Lane | 2012-06-25 15:07:03 | Re: 31.11. Notice Processing - associating notices with calls |