Replication lag is causing issue on one shard

I am running a mongo db sharded cluster with two shards. Each shard has one primary, one arbiter and one secondary. One one shard i am facing issue with replication as it is showing this result

secondset:PRIMARY> db.printSlaveReplicationInfo() source: 192.168.33.154:27017 syncedTo: Wed Dec 03 2014 12:10:51 GMT+0530 (IST) 2644 secs (0.73 hrs) behind the primary It keeps on increasing which si causing issue in migration between shards.

Could you please tell em what can eb the issue and where i can check the cause.