Load balancer based HA for cluster master

[keerthana_k]

Hi,

We are currently running our splunk deployment with a multi-site indexer cluster with search head clustering . The cluster master is present in one site and we would like to configure a stand-by cluster master on the other site for fail-over. We want to implement automated fail-over using a load balancer like HAProxy. Has anyone implemented this? How would you configure your peer indexers and search heads in this case?

Thanks in advance,
Keerthana

[keerthana_k · Accepted solution]

We finally ended up installing a HAProxy load balancer on each of the indexer and search head nodes. All requests to the cluster master was configured to go through the load balancer. We used keepalived libraries in the cluster masters to determine which is the primary and which is standby. Based on this, splunk service will be automatically started or stopped in the cluster masters.

[adonio]

hello there,
from docs:

Although there is currently no master failover capability, you can prepare the indexer cluster for master failure by configuring a stand-by master that you can immediately bring up if the primary master goes down. You can use the same method to replace the master intentionally.

read here:
http://docs.splunk.com/Documentation/Splunk/6.6.1/Indexer/Configurethemaster#Configure_a_stand-by_ma...
will take you right way to here:
http://docs.splunk.com/Documentation/Splunk/6.6.1/Indexer/Handlemasternodefailure
hope it helps

[keerthana_k]

Hi Adonio,

We have set up a stand-by master and we are bringing it up manually now. But in the docs, it is also mentioned that we can do a DNS based or load balancer based failover, so I wanted to know if anyone has actually implemented it.