RO This Run: 87594976 | RO Factor: .94 (94%) RW This Run: 74172726 | RW Factor: .97 (97%)
Drop of 6% in Read, 3% in Write performance on simple sysbench run.
If needed, for details on how sysbench is executed, please pull lp:percona-qa and check performance.sh (and this script can also be used for local performance runs with a bit of tweaking)
Not sure how to proceed with this one. In 5.5, the commit was reviewed thoroughly short of breaking up it to smaller bits, that analysis did not turn up any promising leads, and now we have 200 commits on the top. 5.6 was extensively benchmarked recently. I'm leaning towards marking this as Incomplete.
lpjirasync January 22, 2018 at 10:41 PM
**Comment from Launchpad by: Valerii Kravchuk on: 21-12-2013 15:37:23
Not sure if this affects 5.6 as well, but I see no easy way to check anyway.
**Reported in Launchpad by Roel Van de Paar last update 22-02-2014 04:17:22
http://jenkins.percona.com/view/QA/job/percona-server-5.5-qa-performance/plot/?
Percona-Server-5.5.28-rel29.3-409.Linux.x86_64 - measured performance:
RO This Run: 92549488
RW This Run: 75982562
Percona-Server-5.5.28-rel29.3-410.Linux.x86_64 - measured performance:
RO This Run: 87594976 | RO Factor: .94 (94%)
RW This Run: 74172726 | RW Factor: .97 (97%)
Drop of 6% in Read, 3% in Write performance on simple sysbench run.
If needed, for details on how sysbench is executed, please pull lp:percona-qa and check performance.sh (and this script can also be used for local performance runs with a bit of tweaking)