| From: | Hannu Krosing <hannu(at)2ndQuadrant(dot)com> |
|---|---|
| To: | Simon Riggs <simon(at)2ndQuadrant(dot)com> |
| Cc: | pgsql-hackers(at)postgresql(dot)org, Merlin Moncure <mmoncure(at)gmail(dot)com>, Gregory Stark <stark(at)enterprisedb(dot)com>, Csaba Nagy <nagy(at)ecircle-ag(dot)com>, Heikki Linnakangas <heikki(dot)linnakangas(at)enterprisedb(dot)com>, Robert Haas <robertmhaas(at)gmail(dot)com>, Chris Browne <cbbrowne(at)acm(dot)org>, Jochem van Dieten <jochemd(at)gmail(dot)com>, Dimitri Fontaine <dfontaine(at)hi-media(dot)com>, Richard Huxton <dev(at)archonet(dot)com> |
| Subject: | Re: Transaction Snapshots and Hot Standby |
| Date: | 2008-09-12 11:19:16 |
| Message-ID: | 1221218357.7026.27.camel@huvostro |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Fri, 2008-09-12 at 13:53 +0300, Hannu Krosing wrote:
> 4. Slave keeps copies of removed pages or rows when WAL apply removes
> old versions .
>
> Possible ways to do this
>
> * inside Slave - have some backup store tied to OldestXmin intervals
>
> * variant 1 - have one global store, accessed through shared mem
> * variant 2 - present removed pages to interested backends and
> let them (decide to) keep them
>
> * outside Slave - having file system keep old snapshots as long as
> needed, still must tie to OldestXmin intervals, but most of work done by
> storage layer (SAN or overlay file system).
Possible options for "outside Slave" filesystem snapshooting -
ZFS ( http://en.wikipedia.org/wiki/ZFS ) is very likely usable
Linux LVM + XFS may be usable -
http://arstechnica.com/articles/columns/linux/linux-20041013.ars
Possibly also http://en.wikipedia.org/wiki/Btrfs .
----------------
Hannu
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Gregory Stark | 2008-09-12 11:25:42 | Re: Transaction Snapshots and Hot Standby |
| Previous Message | Simon Riggs | 2008-09-12 11:15:33 | Re: Synchronous Log Shipping Replication |