memcached hung and crashed with: memcached: bucket_engine.c:831: get_engine_handle: Assertion `es' failed.

INFO REPORT <5862.3699.0> 2012-07-02 00:46:16
===============================================================================

memcached<0.3699.0>: memcached: bucket_engine.c:831: get_engine_handle: Assertion `es' failed.

But it took another half-hour for this to get back up into Erlang?:

INFO REPORT <5862.3699.0> 2012-07-02 01:26:58
===============================================================================

ns_log: logging ns_port_server:0:Port server memcached on node 'ns_1@172.16.10.150' exited with status 134. Restarting. Messages: Suspend eq_tapq:replication_ns_1@172.16.10.37 for 5.00 secs
Suspend eq_tapq:replication_ns_1@172.16.10.37 for 5.00 secs
Suspend eq_tapq:replication_ns_1@172.16.10.37 for 5.00 secs
Suspend eq_tapq:replication_ns_1@172.16.10.37 for 5.00 secs
...
memcached: bucket_engine.c:831: get_engine_handle: Assertion `es' failed.

It looks like the system started having trouble gathering stats from both the memcached process and the system-level right before the crash:

CRASH REPORT <5862.3660.0> 2012-07-02 00:46:02
===============================================================================
Crashing process
initial_call

{mb_master,init,['Argument__1']}

pid <5862.3660.0>
registered_name mb_master
error_info
{exit,{timeout,{gen_server,call,[ns_node_disco,nodes_wanted]}},
[

{gen_fsm,terminate,7}

,

{proc_lib,init_p_do_apply,3}

]}
ancestors [ns_server_sup,ns_server_cluster_sup,<5862.51.0>]
messages
[send_heartbeat,send_heartbeat,send_heartbeat,