Hi, Mahesh wrote: > Howver, I find that mysqld reaches > 80% CPU Usage and rsyslog (20% CPU). > The system is a 4 core x86_84 centos system.
1) What are cores? Well you said x86 so you don't run it on an ARM device, but it still can be a Celeron/ATOM processor. MIPS and MHz are still important to know. 2) You need to know how mySQL is working. For example, mySQL will split multiple SESSIONs across the available cores. But each SESSION is bound to one Core. So you can have 64 Cores, but when doing a large INSERT in ONE SESSION you will only utilize 1 of 64 cores with that operation. 3) Your SCHEMA. We don't know anything about it. INDEXes are important. When you have many INDEXes (and triggers) across your table, you have to update multiple things next to the row you will insert itself. 4) And you can do batch INSERTs or you can INSERT each line in a single statement. The different: Doing a batch INSERT will allow you to write the new INDEX only once at the end. But when you INSERT each message in a SINGLE statement mySQL will have to update the INDEX after EVERY INSERT statement. 5) There can be a different between MyISAM and InnoDB. So mySQL (and any other SQL database engine) is probably not the best backend for rsyslog. Maybe that's the reason why log analysis toools (or big data analysis tools in general) won't index into such a backend, instead they user their own. -- Regards, Igor _______________________________________________ rsyslog mailing list http://lists.adiscon.net/mailman/listinfo/rsyslog http://www.rsyslog.com/professional-services/ What's up with rsyslog? Follow https://twitter.com/rgerhards NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE THAT.
