| From: | "Simon Riggs" <simon(at)2ndquadrant(dot)com> |
|---|---|
| To: | "Florian G(dot) Pflug" <fgp(at)phlo(dot)org> |
| Cc: | <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Full page images in WAL & Cache Invalidation |
| Date: | 2007-07-22 22:30:22 |
| Message-ID: | 1185143422.4284.92.camel@ebony.site |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Sun, 2007-07-22 at 17:54 +0200, Florian G. Pflug wrote:
> 1) It seems that the btree code sends out relcache
> invalidation msgs during normal operation
> (No DDL statements are executed). This lets any
> simple flush-all-caches-if-ddl-was-execute scheme
> fail.
Cache invalidation is uniquely associated with these WAL record types:
XLOG_BTREE_NEWROOT
XLOG_BTREE_INSERT_META
XLOG_BTREE_DELETE_PAGE_META
None of those things happen with any real frequency in most use cases.
We don't increase or reduce the number of levels of an index very
frequently.
--
Simon Riggs
EnterpriseDB http://www.enterprisedb.com
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Gregory Stark | 2007-07-22 22:40:51 | Re: Full page images in WAL & Cache Invalidation |
| Previous Message | Simon Riggs | 2007-07-22 22:07:25 | Re: Full page images in WAL & Cache Invalidation |