[rabbitmq-discuss] STOMP Error?

Sat Feb 14 01:49:37 GMT 2009

Okay, I need further clarification on this error...

On Wed, Feb 11, 2009 at 6:51 AM, Tony Garnock-Jones <tonyg at lshift.net> wrote:
> Darien Kindlund wrote:
>> =ERROR REPORT==== 5-Feb-2009::13:05:28 ===
>> STOMP error frame sent:
>> Message: "Processing error"
>> Detail: "{timeout,{gen_server,call,[rabbit_persister,serial]}}\n"
>
> This looks like the persister rollover problem I mentioned just a moment
> ago: when there's a large volume of data, it can take a while for the
> log file to roll over. Processes calling the persister during a long
> rollover may block long enough for their timeout to fire. A fix is in
> the works!

Are you referring to rolling the persister log file?  To be clear,
when things are working "normally", this activity is normally logged
as:

=INFO REPORT==== 8-Feb-2009::06:34:23 ===
Rolling persister log to
"/var/lib/rabbitmq/mnesia/rabbit/rabbit_persister.LOG.previous"

=INFO REPORT==== 8-Feb-2009::06:39:12 ===
Rolling persister log to
"/var/lib/rabbitmq/mnesia/rabbit/rabbit_persister.LOG.previous"

=INFO REPORT==== 8-Feb-2009::06:44:02 ===
Rolling persister log to
"/var/lib/rabbitmq/mnesia/rabbit/rabbit_persister.LOG.previous"

... right?

1) Is it "normal" for this activity to occur every 5 minutes?  Is this
a fixed time or is it variable?  If it's variable, what factors
control when and how often this occurs?

2) When one STOMP connection generates this error, does it affect
any/all future STOMP connections?  Meaning, is this error more likely
to occur with 500 connections as opposed to 5?  Or does that not
matter?

Thanks,
-- Darien