CODEWITHSUNDEEP
ejabberd
sanket Thakare
sanket Thakare

Reputation: 115

ejabberd Clustering with master Slave and Master Master not avoid single point of failure

Whenever Master gets down in ejabberd Cluster slave will not be able to process request. i follow the tutorial from : http://chadillac.tumblr.com/post/35967173942/easy-ejabberd-clustering-guide-mnesia-mysql and try with join_node() and join_as_master() also. But its not working. My cluster is working and i used mysql database as separate database for both node. I tried also with join as master slave and master master method. Both not working to avoid single point of fauilure. Is it necessary to synchronize slave and master mnesia database? please any one can help me in this...

Upvotes: 1

Views: 1041

Answers (1)

Mickaël Rémond
Mickaël Rémond

Reputation: 9055

There is no single point of failure with ejabberd if you configure it the right way. There is no such thing as master-slave in ejabberd. ejabberd is master-master and if you configure it with join_clustercommand as described in official ejabberd documentation: http://docs.ejabberd.im/admin/guide/clustering/

You need to configure the other services as redundant as well: - Load balancer - MySQL

Once done, I do not see what you can see as point of failure. If a node has to go down, the service keeps on working fine for users connected on other nodes and the one that have been disconnected can reconnect just fine on any of the remaining node.

If you means that if you stop both nodes and then try to start only the first stopped, this is normal behaviour of Mnesia distributed database. It is then a duplicate question already answered here: ejabberd clustering, Slave doesn't work when master goes down

Note: The blog post you are pointing to is obsolete.

Upvotes: 1

Related Questions