host:11210 port get down in case of only access.log.old (corrupted one)

Steps to Reproduce :-
1) Load 10,000 Items on two node cluster.
2) Run access scanner using cbepctl.Wait for access.log to get created
3) Load another set of Items and wait for access.log.old to get created
4) corrupt the access.log.old and delete the access.log file
5) Stop and Start the server.
6) run cbstats to collect the raw warmup info but we receive following error
./cbstats 10.3.121.103:11210 raw warmup
Traceback (most recent call last):
File "./cbstats", line 398, in <module>
main()
File "./cbstats", line 395, in main
c.execute()
File "/home/rohit/Official/Views/ep-engine/management/clitool.py", line 58, in execute
f[0](mc, *args[2:], **opts._dict_)
File "./cbstats", line 46, in g
f(*args[:n])
File "./cbstats", line 255, in stats_raw
stats_formatter(mc.stats(arg))
File "/home/rohit/Official/Views/ep-engine/management/mc_bin_client.py", line 405, in stats
cmd, opaque, cas, klen, extralen, data = self._handleKeyedResponse(None)
File "/home/rohit/Official/Views/ep-engine/management/mc_bin_client.py", line 92, in _handleKeyedResponse
cmd, errcode, opaque, cas, keylen, extralen, rv = self._recvMsg()
File "/home/rohit/Official/Views/ep-engine/management/mc_bin_client.py", line 74, in _recvMsg
raise exceptions.EOFError("Got empty data (remote died?).")
EOFError: Got empty data (remote died?).

Could not connect to 10.3.121.103:11210: Connection refused
While 11211 port is still running
rohit@rohit-HP-ProBook-6460b:~/Official/Views/ep-engine/management$ ./cbstats 10.3.121.103:11211 raw warmup
ep_warmup: enabled
ep_warmup_dups: 0
ep_warmup_estimated_item_count: unknown
ep_warmup_estimated_warmup_count: unknown
ep_warmup_key_count: 0
ep_warmup_keys_time: 900
ep_warmup_min_item_threshold: 100
ep_warmup_min_memory_threshold: 100
ep_warmup_oom: 0
ep_warmup_state: done
ep_warmup_thread: complete
ep_warmup_time: 1041
ep_warmup_value_count: 0

Please see attachment for ns_server, diag etc. log files