Hello,

I've configured a Sharding Cluster with below components.

3 config serves
2 mongos
2 shards (3 replica members each)

Case 1: When 1 out of 3 config server is down. As balancer/split chunks doesn't work all the data is being written to primary shard. Later when the conifg server is back. Data is not being split to other shards.

Is there a way that after config server is back that balancer/split chunks migrate to other shards?

Case 2: As My primary server in mongo is listening to 27011 port. After failover the new primary is 27012. But as the mongos is configured to port 27011. It is failing to connect to shard.

How to configure mongos to use the primary member of replset.?