Percona Cluster Master-Master (Slave Node) Dynamic Config Generation
It is based on Percona Docker Image
- Copy from a defined TAR (url) the template my.cnf and cluster.cnf that you wish to run on the node
- Through a folder, sync/replicated to each node (nfs/gluster/ocfs/etc) keeps a heartbeat of nodes
that are UP and running with the heartbeat.
Its able to keep up-to-date the cluster ips of a galera cluster.
It runs supervisor for both heartbeat and mysql daemon
Big cluster, docker, many containers, many slaves/donors that come up & down on-demand.
How you keep the cluster.cnf updated ? Could come with thousand of ideas, but i think this keeps up by itself
and its quite simple to extend.
Contributions / Forks are welcome :)