https://www.percona.com/blog/scaling-the-percona-kubernetes-operator-for-percona-xtradb-cluster/ Fri, 10 Jul 2020 15:13:20 +0000 hourly 1 https://wordpress.org/?v=6.5.2 https://www.percona.com/blog/scaling-the-percona-kubernetes-operator-for-percona-xtradb-cluster/#comment-10972465 Thu, 09 Jul 2020 16:53:55 +0000 https://www.percona.com/blog/?p=69714#comment-10972465 Correct, removing AntiAffinity is only tolerable for lower environments like QA or internal Dev envs. For production environments not having a node per pod will defeat the whole purpose of using Percona XtraDB Cluster which is High Availability

]]> https://www.percona.com/blog/scaling-the-percona-kubernetes-operator-for-percona-xtradb-cluster/#comment-10972464 Thu, 09 Jul 2020 16:20:14 +0000 https://www.percona.com/blog/?p=69714#comment-10972464 I think it is worth to note if you change AntiAffinity you’re increasing risks for cluster availability. It may be good for testing but for production workloads you surely want to run nodes on separate hosts, or otherwise in a way concurrent failure of multiple pods is unlikely.

]]>