Hi, we are doing concurrent read and write at time to a table whche is having about100,000 records.
The cluster has 3 nodes percona cluster with Haproxy LB.
At a time concurrently about 20 request goes to the table ( read and write), So that the mySQL connection goes to about 600 and drive the entire system to slowness because of deadlocks.
Do you advice any parameter to change in the my.cfn file to solve this.
Hello, thanks for posting. There’s maybe not enough to go on here.
There are quite a lot of blog posts on the Database Performance blog and documentation that look at the issue of deadlocks. Some links:
[url]Frequently Asked Questions - Percona XtraDB Cluster
[url]https://www.percona.com/blog/2014/09/11/openstack-users-shed-light-on-percona-xtradb-cluster-deadlock-issues/[/url]
[url]https://www.percona.com/blog/2012/08/17/percona-xtradb-cluster-multi-node-writing-and-unexpected-deadlocks/[/url]
If you still need help though, can I get you to summarize details of your environment please as that will help the engineers to help you more efficiently?
[LIST]
[]Percona XtraDB Cluster version
[]Software platform & version
[]Any error log file entries being written?
[]Configuration file (for each node)
[]A description of the topology, environments
[]Whether this is a new (development) or it’s a new problem in an existing system
[/LIST] Thanks!