AnsweredAssumed Answered

Database Clustering Best Practices?

Question asked by jkearney on Mar 2, 2016

Good Afternoon,


I am looking for information on what the best practices are for clustering databases for sugarcrm?

Currently we have a MariaDB Galera cluster setup with 2 HAproxy load balancers taking care of routing traffic from our Virtual IP address(I've attached a rough chart because I think I explained that badly, I followed this to set it up mostly Making HAProxy High Available for MySQL Galera Cluster | FromDual - MySQL, Galera Cluster, MariaDB and Percona Server su… ). This seems to be working, and I love that the proxy is cluster state aware so failover is all automated. But I have been told that this is not the proper setup for a sugarcrm environment by some coworkers, and my research hasn't yielded anything that really says definitively one way or the other.  I've seen some web pages talk about async replication as a way to 'cluster' the databases ( ) but it didn't say not to use a synchronous cluster specifically. And I found one that talks about using Galera specifically (Clustering SugarCRM with MySQL Galera for High Availability and Performance | Severalnines ). So I'm just curious what other people are using to make their databases highly available? Does anyone have any reasons why we shouldn't be using a synchronous cluster for our sugarcrm databases? I'm very new to this and would love any information that you can provide! Thank you in advance!