[rabbitmq-discuss] memory leak rabbitmq 3.1.5

Santi Nuñez-Cacho brian.santiago at gmail.com
Thu Sep 26 13:51:40 BST 2013


I expect 0 stomp connetions. The unique one connection on 61613 should be
the cluster one.

[root at rabbit03.bcn ~]#  rabbitmqctl eval
'length(supervisor:which_children(rabbit_stomp_client_sup_sup)).'
3802
...done.
[root at rabbit03.bcn ~]# lsof -i :61613
COMMAND    PID     USER   FD   TYPE  DEVICE SIZE/OFF NODE NAME
ruby      5638     root    6u  IPv4   25097      0t0  TCP
rabbit03.bcn.softonic.lan:59851->rabbit:61613 (ESTABLISHED)
beam.smp 27183 rabbitmq   21u  IPv6 4521334      0t0  TCP *:61613 (LISTEN)
[root at rabbit03.bcn ~]# netstat -apn | grep 61613
tcp        0      0 192.168.250.135:59851       192.168.255.12:61613
 ESTABLISHED 5638/ruby
tcp        0      0 :::61613                    :::*
 LISTEN      27183/beam.smp
[root at rabbit03.bcn ~]#


On Thu, Sep 26, 2013 at 2:23 PM, Simon MacMullen <simon at rabbitmq.com> wrote:

> Please keep rabbitmq-discuss on CC.
>
> So that shows 30 STOMP connection processes. Is that what you would
> expect? Does that correspond with what you see in the management connection
> listing? What about "lsof -i :61613"?
>
> Cheers, Simon
>
>
> On 26/09/2013 12:14PM, Santi Nuñez-Cacho wrote:
>
>> This is the response:
>>
>> [root@ ~]#  rabbitmqctl eval
>> 'length(supervisor:which_**children(rabbit_stomp_client_**sup_sup)).'
>> 30
>> ...done.
>> [root@ ~]#
>>
>>
>> thank you very much
>>
>>
>> On Thu, Sep 26, 2013 at 12:03 PM, Simon MacMullen <simon at rabbitmq.com
>> <mailto:simon at rabbitmq.com>> wrote:
>>
>>     On 26/09/2013 10:15AM, Santi Nuñez-Cacho wrote:
>>
>>         I start disabling one by one the plugins and I come across with
>>         the most
>>         feasible candidate. Stomp.
>>
>>         I disable stomp plugin and memory settle.
>>
>>
>>     Hmm. Odd. I'm certainly not aware of a memory leak in STOMP.
>>
>>     Are you sure nothing is opening STOMP connections to your broker?
>>     Connections will only appear in mgmt when a CONNECT frame has been
>>     sent, so if something is just opening TCP sockets then they will not
>>     show up.
>>
>>     What does
>>     $ rabbitmqctl eval
>>     'length(supervisor:which___**children(rabbit_stomp_client__**
>> _sup_sup)).'
>>
>>
>>     say when the system is in this state?
>>
>>     Cheers, Simon
>>
>>     --
>>     Simon MacMullen
>>     RabbitMQ, Pivotal
>>
>>
>>
> --
> Simon MacMullen
> RabbitMQ, Pivotal
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rabbitmq.com/pipermail/rabbitmq-discuss/attachments/20130926/06f6063b/attachment.htm>


More information about the rabbitmq-discuss mailing list