| From: | Stephen Frost <sfrost(at)snowman(dot)net> |
|---|---|
| To: | Torsten Zuehlsdorff <mailinglists(at)toco-domains(dot)de> |
| Cc: | Vladimir Borodin <root(at)simply(dot)name>, Pavel Stehule <pavel(dot)stehule(at)gmail(dot)com>, Dinesh Chandra 12108 <Dinesh(dot)Chandra(at)cyient(dot)com>, "Madusudanan(dot)B(dot)N" <b(dot)n(dot)madusudanan(at)gmail(dot)com>, "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: Backup taking long time !!! |
| Date: | 2017-01-23 17:20:54 |
| Message-ID: | 20170123172054.GF18360@tamriel.snowman.net |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Greetings,
* Torsten Zuehlsdorff (mailinglists(at)toco-domains(dot)de) wrote:
> I just have around 11 TB but switched to ZFS based backups only. I'm
> using snapshots therefore which gives some flexibility. I can
> rolback them, i can just clone it and work with a full copy as a
> different cluster (and just the differences are stored) and i can
> send them incrementally to other servers. This is very fine for my
> use cases but it doesn't fit everything of course.
While that's another approach, it does require that those snapshots are
performed correctly (both by you and by ZFS) and are entirely atomic to
the entire PG instance.
For example, I don't believe ZFS snapshots will be atomic if multiple
ZFS filesystems on independent ZFS pools are being used underneath a
single PG instance.
And, as others have also said, always test, test, test.
Thanks!
Stephen
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Stephen Frost | 2017-01-23 17:23:01 | Re: Backup taking long time !!! |
| Previous Message | Jeff Janes | 2017-01-23 17:12:22 | Re: Backup taking long time !!! |