ofbiz-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nicolas Malin <nicolas.ma...@nereide.fr>
Subject Re: Async Service Thread Locking
Date Wed, 01 Feb 2017 19:43:31 GMT
Hello Ryan;

Yes 10.04 it's old.

 From my experience :

* check you crashed service : when you restart your OFBiz, if you have 
some scheduled service, the "crash"  duplicate them

* check you java heap space with jmx because when a thread overload 
ressources or something like that you have an unexpected result on job 
manager

* prefer to use more than one pool if you have too many async service.

* Warning on huge async service. On a big project I down the jobs 
authorize on running at only 2. I prefer load the queue instead of 
overload  the server. And if you service queue is too high, deploy a new 
ofbiz to help this pool ;)

Nicolas


Le 01/02/2017 à 18:39, Ryan Moriarty a écrit :
> This is OFBiz 10.04. Quite old, I know, but I'm stuck for the moment.
>
> I haven't seen any errors about being able to queue jobs, but I may be
> searching for the wrong text
>
> The settings are as they are when I inherited the project, I can't speak to
> why they are as they are.
>
> On Tue, Jan 31, 2017 at 10:25 PM, Jacques Le Roux <
> jacques.le.roux@les7arts.com> wrote:
>
>> Le 31/01/2017 à 21:24, Ryan Moriarty a écrit :
>>
>>> I'm having difficulty with asynchronous jobs not completing. Many jobs
>>> finish normally, but most show "Running" into perpetuity. I see no pattern
>>> in services that finish and services that continue running indefinitely. I
>>> have five invoker threads that appear to always be sleeping. We're using a
>>> single-server config. This is my thread pool config:
>>>
>>>    <thread-pool send-to-pool="pool"
>>>                        purge-job-days="4"
>>>                        failed-retry-min="3"
>>>                        ttl="18000000"
>>>                        wait-millis="750"
>>>                        jobs="10"
>>>                        min-threads="5"
>>>                        max-threads="15"
>>>                        poll-enabled="true"
>>>                        poll-db-millis="20000">
>>>               <run-from-pool name="pool"/>
>>>
>>> I saw a message that suggested I would be out of heap space, but my max
>>> heap size is 5280M, and we have only ~100 concurrent users, so that
>>> doesn't
>>> seem like an issue. Could it be a deadlock? How would I resolve that? Any
>>> suggestions are welcome.
>>>
>>> Ryan Moriarty
>>> Software Developer
>>> Able Aerospace Services
>>>
>>>
>> Hi Ryan,
>>
>> Which OFBiz version are you using?
>>
>> With jobs=10, don't you ever get "Unable to queue job" exceptions?
>>
>> And why did you change the default setting at large?
>>
>> Jacques
>>
>>


Mime
View raw message