continuous load causes memcached server errors

1) setup 1 node cluster
2) set ram size to 1570MB
3) create bucket with 1570MB
4) load 200k keys with valuesize if 1024 in a loop (200 MB)
after while we start getting server errors (SERVER_ERROR proxy downstream closed 127.0.0.1), also getting errors when trying to set directly to memcached:11210 (EOFError: Got empty data (remote died?))
the server shows up as pending in the web gui and stats against memcached returns nothing

Eventually it seems to somewhat recover, I can get stats against memcached, and set items through moxi. However trying to get stats through moxi fails.

GUI log (I never manually initiated a bucket shutdown)

Bucket "default" loaded on node 'ns_1@127.0.0.1' in 0 seconds. (repeated 1 times)
Shutting down bucket "default" on 'ns_1@127.0.0.1' for server shutdown
Control connection to memcached on 'ns_1@127.0.0.1' disconnected: {{badmatch,
{error,
timeout}},
[

{mc_client_binary, stats_recv, 4}

,

{mc_client_binary, stats,4}

,

{ns_memcached, handle_call, 3}

,

{gen_server, handle_msg, 5}

,

{proc_lib, init_p_do_apply, 3}

]}
Bucket "default" loaded on node 'ns_1@127.0.0.1' in 0 seconds.