[rabbitmq-discuss] Update: [Bug] Memory leak in 3.1.3 (Ubuntu)
Louis Bouchard
louis.bouchard at canonical.com
Thu Jul 18 10:17:48 BST 2013
Hello,
Le 16/07/2013 15:16, Matthias Radestock a écrit :
> If you are planning on running this test for a while, I suggest you take
> a reading of the memory values reported by 'rabbitmqctl status' near the
> beginning and end of the run, since that will help pinpoint where the
> memory is going.
>
> Regards,
>
> Matthias.
My instance has now been running for almost two days. Here is some data :
saucy> sudo rabbitmqctl status
Status of node 'rabbit at server-656f503b-0965-4e01-94d9-58b177bbec3d' ...
[{pid,2150},
{running_applications,[{rabbit,"RabbitMQ","3.1.3"},
{mnesia,"MNESIA CXC 138 12","4.9"},
{os_mon,"CPO CXC 138 46","2.2.12"},
{xmerl,"XML parser","1.3.3"},
{sasl,"SASL CXC 138 11","2.3.2"},
{stdlib,"ERTS CXC 138 10","1.19.2"},
{kernel,"ERTS CXC 138 10","2.16.2"}]},
{os,{unix,linux}},
{erlang_version,"Erlang R16B01 (erts-5.10.2) [source] [64-bit]
[async-threads:30] [kernel-poll:true]\n"},
{memory,[{total,36267272},
{connection_procs,82704},
{queue_procs,60600},
{plugins,0},
{other_proc,13558976},
{mnesia,59392},
{mgmt_db,0},
{msg_index,39616},
{other_ets,744408},
{binary,1335160},
{code,16368314},
{atom,594537},
{other_system,3423565}]},
{vm_memory_high_watermark,0.4},
{vm_memory_limit,840009318},
{disk_free_limit,1000000000},
{disk_free,8495063040},
{file_descriptors,[{total_limit,924},
{total_used,4},
{sockets_limit,829},
{sockets_used,2}]},
{processes,[{limit,1048576},{used,131}]},
{run_queue,0},
{uptime,171822}]
...done.
saucy> free
total used free shared buffers cached
Mem: 2050804 1243784 807020 0 82532 979120
-/+ buffers/cache: 182132 1868672
Swap: 0 0 0
saucy> ps auxww | grep beam
rabbitmq 2150 3.3 1.8 336912 38340 ? Sl Jul16 96:28
/usr/lib/erlang/erts-5.10.2/bin/beam -W w -K true -A30 -P 1048576 --
-root /usr/lib/erlang -progname erl -- -home /var/lib/rabbitmq -- ...
So there is clearly some memory leakage here. It may not come from
rabbitmq (which does not seem to be consuming all that memory) but it is
clearly triggered by it.
During my previous test, restarting rabbitmq did not release the used
memory so it might also be some kind of kernel leak triggered by rabbitmq.
I'll let the instance run for now.
Kind regards,
...Louis
--
Louis Bouchard
Backline Support Analyst
Canonical Ltd
Ubuntu support: http://canonical.com/support
More information about the rabbitmq-discuss
mailing list