> > Mr. Bacula Users, > > > > What do you think of those MySQL tuning hints for Bacula, specifically > > about this innodb_flush_log_at_trx_commit option? > > => > > http://www.innovation-brigade.com/index.php?module=Content&type=user&func=display&tid=1&pid=2 > > it seems to improve MySQL attributes despooling time by a 1000 factor, > > from hundreds transactions per second to millions. > > Truth is, innodb_flush_log_at_trx_commit = 2 should be the default > option. innodb_flush_log_at_trx_commit = 1 is going to be slow whenever > there is a very high rate of small commits, because every commit causes > a flush to disk. Setting it to 1 should only be done when you literally > cannot afford to lose a single commit. Setting it to 2, all you are > risking is the *possible* loss of *up to* the last second of commits, > *only* in the case of a failure of the storage layer, an unexpected loss > of power, or a complete crash of the operating system. And in any one > of those situations, you're going to have to completely re-run your > backups from scratch anyway.
Thanks Phil! That was very enlightening. I think for Bacula type of workload is a very important MySQL setting to change. Do you have any information if for Postgresql using the synchronous_commit = on option would have similar effects? Heitor. > ------------------------------------------------------------------------------ Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server from Actuate! Instantly Supercharge Your Business Reports and Dashboards with Interactivity, Sharing, Native Excel Exports, App Integration & more Get technology previously reserved for billion-dollar corporations, FREE http://pubads.g.doubleclick.net/gampad/clk?id=164703151&iu=/4140/ostg.clktrk _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
