Description of problem: If a slave node goes down, the geo-rep session connecting to that particular node from master doesn't go to faulty immediately,but it take 15 to 20 min to go to faulty and consequently it doesn't sync any files to the slave from that brick or subvolume on master. 
And once that node comes back, the connection will be established, but it does first xsync crawl. 

Version-Release number of selected component (if applicable):glusterfs-3.4.0.30rhs-2.el6rhs.x86_64


How reproducible: Happens everytime


Steps to Reproduce:
1.Create and start a geo-rep relationship between master and slave.
2.Bring down one of the node from slave.
3.create some files on the slave.
4. Check if those files are synced to slave or not / 

Actual results: If a slave machines goes down, that particular session doesn't 
go down immediately.

Expected results: It should goes down immediately and also shouldn't it should pick another sane slave machine to sync the data . 


Additional info: