Re: [REVIEW] Re: Compression of full-page-writes

<p dir="ltr">That's not entirely true. CRC-32C beats pretty much everything with the same length quality-wise and has
bothhardware implementations and highly optimized software versions.<div class="gmail_quote">Em 12/09/2014 17:18, "Ants
Aasma"<<a href="mailto:ants@cybertec.at">ants@cybertec.at</a>> escreveu:<br type="attribution" /><blockquote
class="gmail_quote"style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On Fri, Sep 12, 2014 at 10:38
PM,Heikki Linnakangas<br /> <<a href="mailto:hlinnakangas@vmware.com">hlinnakangas@vmware.com</a>> wrote:<br />
>I don't mean that we should abandon this patch - compression makes the WAL<br /> > smaller which has all kinds
ofother benefits, even if it makes the raw TPS<br /> > throughput of the system worse. But I'm just saying that
theseTPS<br /> > comparisons should be taken with a grain of salt. We probably should<br /> > consider switching
toa faster CRC algorithm again, regardless of what we do<br /> > with compression.<br /><br /> CRC is a pretty
awfullyslow algorithm for checksums. We should<br /> consider switching it out for something more modern. CityHash,<br
/>MurmurHash3 and xxhash look like pretty good candidates, being around<br /> an order of magnitude faster than CRC.
I'mhoping to investigate<br /> substituting the WAL checksum algorithm 9.5.<br /><br /> Given the room for improvement
inthis area I think it would make<br /> sense to just short-circuit the CRC calculations for testing this<br /> patch
tosee if the performance improvement is due to less data being<br /> checksummed.<br /><br /> Regards,<br /> Ants
Aasma<br/> --<br /> Cybertec Schönig & Schönig GmbH<br /> Gröhrmühlgasse 26<br /> A-2700 Wiener Neustadt<br /> Web:
<ahref="http://www.postgresql-support.de" target="_blank">http://www.postgresql-support.de</a><br /><br /><br /> --<br
/>Sent via pgsql-hackers mailing list (<a
href="mailto:pgsql-hackers@postgresql.org">pgsql-hackers@postgresql.org</a>)<br/> To make changes to your
subscription:<br/><a href="http://www.postgresql.org/mailpref/pgsql-hackers"
target="_blank">http://www.postgresql.org/mailpref/pgsql-hackers</a><br/></blockquote></div>