Bug 1105543

Description of problem:
=======================
glusterd hangs when a node with stale snap entries are attached to the cluster


Version-Release number of selected component (if applicable):
=============================================================

glusterfs 3.6.0.13

How reproducible:
================
1/1

Steps to Reproduce:
===================

4 node cluster : Node1 Node2 Node3 Node4 

Attach a node(Node5) to the cluster while snapshot creation (snap_vol0_1.. snap_vol0_n) is in progress.
Check snapshots on the newly added peer. 


On node 4 :
----------
After snapshots are completed, detach the node and attach it again while starting snapshot creation again(snap1.. snapn)


Now probe is successful, but gluster peer status hangs as it is trying to start bricks of the stale snap entries (snap_vol0_1) which is not present on Node4.

 E [glusterd-handshake.c:85:get_snap_volname_and_volinfo] 0-management: Failed to fetch s
nap snap_vol0_1
[2014-06-06 10:49:52.132065] E [glusterd-handshake.c:196:build_volfile_path] 0-management: Failed to get snap volinfo
 from path (/snaps/snap_vol0_1/a94839bbaf994733a5b591bb59731d9c.snapshot16.lab.eng.blr.redhat.com.var-run-gluster-sna
ps-a94839bbaf994733a5b591bb59731d9c-brick4-b0)


Also checking gluster peer status on the newly added  peer shows the below status : 

[root@snapshot01 ~]# gluster peer status
Number of Peers: 1

Hostname: 10.70.40.172
Uuid: 2c797de6-e1e0-4a9a-8729-f5d3ce2de2a1
State: Sent and Received peer request (Connected)

Its shows the above status of one Node and the status other nodes are not listed . 

Actual results:


Expected results:


Additional info: