Cluster goes into bad state after auto failover with memcached buckets in the cluster

1. Create an .ini file with atleast 4 servers with latest CB in it. (example given below)
2. Clone git://github.com/bharath-gp/testrunner.git and checkout autofailovertests
3. Run the following tests using the ini file created above
   ./testrunner i centos_x6401_01-autofailover_upr.ini  -t failover.AutoFailoverTests.AutoFailoverTests.test_autofailover,timeout=5,num_node_failures=1,memcached_buckets=2,failover_action=stop_server,nodes_init=3,num_items=1000

The test passes. But the state of the cluster goes real bad. Cannot access the cluster after the test using UI or subsequent tests cannot create any buckets in the cluster. Any cluster management calls returns  ["Unexpected server error, request logged."]as status. This is blocking any subsequent tests run in our daily run.

Attaching the logs of the cluster after the test has run. http://qa.sc.couchbase.com/job/cen006-nserv-autofailover-buckets/19/artifact/logs/testrunner-17-Mar-30_08-15-20/test_12/