I find "The timeout callback method for a programmatically created
persistent or non-persistent timer will be invoked on the JVM on which
the timer was created." to be a bit too restrictive.
I believe nobody needs to have the timer be executed on the same JVM and
implementations might want to take load balancing into consideration.
How about: "The timeout callback method for a programmatically created
persistent or non-persistent timer will be invoked on the JVM on which
the timer was created or if the container is distributed any other JVM
where the same container is active."?
Carlo
On 05/03/2012 06:23 PM, Marina Vatkina wrote:
> I had added this text at the bottom of the p.322 in the EDR (as part
> of various clarifications):
>
> "The timeout callback method for a programmatically created persistent
> or non-persistent timer will be invoked on the JVM on which the timer
> was created. In the event of a container crash or container shutdown,
> the timeout callback method for a programmatically created persistent
> timer that has not been cancelled will be invoked on a new JVM when
> the container is restarted or on another JVM instance across which the
> container is distributed."
>
> If you think it needs to be expanded, please suggest your version or
> the extra text.
>
> thanks,
> -marina
>
> Reza Rahman wrote:
>> +1
>>
>> On 5/3/2012 1:33 AM, Adam Bien wrote:
>>> Hi *,
>>>
>>> given persistent timers are coordinated in a shared DB, they should
>>> work-out-of the box in a cluster. It means: a timer should fire on a
>>> node, but not all nodes at the same time.
>>>
>>> It is one of the FAQs. Should we clarify that (or the supposed
>>> behavior) in the spec?
>>>
>>> thanks!,
>>>
>>> adam
>>>
>>> -----
>>> No virus found in this message.
>>> Checked by AVG - www.avg.com
>>> Version: 2012.0.1913 / Virus Database: 2425/4974 - Release Date:
>>> 05/02/12
>>>
>>>
>>>
>>