From:
Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To:
Florian Weimer <fw(at)deneb(dot)enyo(dot)de>
Cc:
"Simon Riggs" <simon(at)2ndquadrant(dot)com>, pgsql-hackers(at)postgresql(dot)org,
pgsql-patches(at)postgresql(dot)org
Subject:
Re: [HACKERS] wal_checksum = on (default) | off
Date:
2007-01-04 20:22:10
Message-ID:
26352.1167942130@sss.pgh.pa.us (view raw or flat )
Thread:
2007-01-04 14:27:56 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-04 15:00:56 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-04 15:56:39 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-04 16:09:59 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-04 16:51:50 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-04 17:13:57 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-04 17:53:44 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-05 10:01:31 from "Zeugswetter Andreas ADI SD" <ZeugswetterA(at)spardat(dot)at>
2007-01-05 10:35:47 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-05 11:30:46 from "Zeugswetter Andreas ADI SD" <ZeugswetterA(at)spardat(dot)at>
2007-01-05 14:39:53 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-05 16:25:08 from "Zeugswetter Andreas ADI SD" <ZeugswetterA(at)spardat(dot)at>
2007-01-05 21:41:01 from Jim Nasby <decibel(at)decibel(dot)org>
2007-01-06 03:57:32 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-06 04:05:56 from "Joshua D(dot) Drake" <jd(at)commandprompt(dot)com>
2007-01-06 14:41:02 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-06 21:30:08 from Bruce Momjian <bruce(at)momjian(dot)us>
2007-01-11 04:32:51 from Bruce Momjian <bruce(at)momjian(dot)us>
2007-01-11 11:32:12 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-11 14:01:09 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-11 14:54:52 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-11 14:59:26 from Gregory Stark <stark(at)enterprisedb(dot)com>
2007-01-11 15:07:13 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-11 16:18:36 from Gregory Stark <stark(at)enterprisedb(dot)com>
2007-01-11 16:22:19 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-11 16:25:28 from Gregory Stark <stark(at)enterprisedb(dot)com>
2007-01-11 16:37:38 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-11 17:06:11 from Gregory Stark <stark(at)enterprisedb(dot)com>
2007-01-11 23:10:38 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-12 10:09:28 from Martijn van Oosterhout <kleptog(at)svana(dot)org>
2007-01-04 16:58:51 from Florian Weimer <fweimer(at)bfk(dot)de>
2007-01-04 17:05:04 from "Simon Riggs" <simon(at)2ndquadrant(dot)com>
2007-01-04 17:07:03 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-04 19:18:12 from Ron Mayer <rm_pg(at)cheapcomplexdevices(dot)com>
2007-01-04 19:43:11 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-04 20:16:02 from Florian Weimer <fw(at)deneb(dot)enyo(dot)de>
2007-01-04 20:22:10 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
2007-01-04 21:48:56 from Florian Weimer <fw(at)deneb(dot)enyo(dot)de>
2007-01-04 22:03:45 from Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Lists:
pgsql-hackers pgsql-patches
Florian Weimer <fw(at)deneb(dot)enyo(dot)de> writes:
> * Tom Lane:
>> There's a lot of math behind CRCs but AFAIR Adler's method is pretty
>> much ad-hoc.
> Correct me if I'm wrong, but the main reason for the WAL CRC is to
> detect partial WAL writes (due to improper caching, for instance).
Well, that's *a* reason, but not the only one, and IMHO not one that
gives any particular guidance on what kind of checksum to use.
> This means that you're out of the realm of traditional CRC analysis
> anyway, because the things you are guarding against are neither burts
> errors nor n-bit errors (for small n).
I think short burst errors are fairly likely: the kind of scenario I'm
worried about is a wild store corrupting a word of a WAL entry while
it's waiting around to be written in the WAL buffers. So the CRC math
does give me some comfort that that'll be detected.
regards, tom lane
In response to
Responses
pgsql-hackers by date
Next :From: Gurjeet SinghDate: 2007-01-04 20:26:12Subject : Re: [HACKERS] [Fwd: Index Advisor]Previous :From : Florian WeimerDate : 2007-01-04 20:16:02Subject : Re: [HACKERS] wal_checksum = on (default) | off
pgsql-patches by date
Next :From: Gurjeet SinghDate: 2007-01-04 20:26:12Subject : Re: [HACKERS] [Fwd: Index Advisor]Previous :From : Florian WeimerDate : 2007-01-04 20:16:02Subject : Re: [HACKERS] wal_checksum = on (default) | off
