Hi,

We have implemented a replica set which uses 2 nodes (say nodeA and nodeB) for data and another node for arbiter. Mongo version used is 2.4.2 on Centos 6 x64

The system takes around 30 seconds or more to failover and elect a Primary in case any one member which was Primary goes down. We have scenarios where the network can fluctuate and recover on its own.

Is there a way to increase or decrease this failover detection time ? Any configuration parameters ?

Thanks

Omer