tomee-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Leonardo K. Shikida" <shik...@gmail.com>
Subject Re: Cluster + @Schedule
Date Sat, 18 Oct 2014 13:37:44 GMT
Hi Romain

I think... I am getting the idea :-)

>From your answer, I infer that if a have a bean with some method annotated
with @Timeout then "task" = execute this method.

And each "task" has a single "trigger". So, if I try to create a new
"trigger" for this "task", I will end up just replacing the existent
"trigger" (or if it has the same cron expression, just being ignored).
Persistent, of course.

I thought that if I create 3 triggers for the same task to be run at the
same time, then the task would be executed 3 times (one for each trigger).

I really don't know what the spec talks about this (if it talks about it).

What's fancy about this IMO is that this 1-1 relationship between triggers
and beans is clear when we're creating timers statically, using annotations
(@Schedule), but it wasn't clear for me for dynamically created timers
(Timer created = timerService.createCalendarTimer(exp,config))



[]

Leo

On Sat, Oct 18, 2014 at 10:23 AM, Romain Manni-Bucau <rmannibucau@gmail.com>
wrote:

> Le 18 oct. 2014 14:47, "Leonardo K. Shikida" <shikida@gmail.com> a écrit :
> >
> > > "hello 1413583260198" was created in a previous attempt, so my feeling
> is
> >
> > > that, somehow, even creating a new timer on every server restart, tomEE
> (or
> > > Quartz) is assuming that the calendar timer is already created
> (checking by
> > > cron expression maybe?). In fact, in the DB, there are only 3 triggers
> > > everytime.
> > > >
> > > >
> > >
> > > If persistent that s possible
> > >
> > >
> > Hi Romain
> >
> > Yeah, but then is it a bug? I mean, if I am creating a new timer, even if
> > it has the same calendar cron expression, it should not be silently
> ignored
> > just because there's already one in the DB, right?
> >
>
> If that s the same for the same bean with same data it doesnt shock me. If
> not persistence adapter id maybe needs some tweak
>
> > Yes, it's persistent. It has to be to make quartz cluster work.
> >
> > */5 no?
> > >
> > >
> > sorry, I feel so stupid about this :-)
> >
> > TIA
> >
> > Leo
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message