StackOverflow2013

Note that there are some explanatory texts on larger screens.

plurals

POquartz: preventing concurrent instances of a job in jobs.xml
text
Body
copied!<p>This should be really easy. I'm using Quartz running under Apache Tomcat 6.0.18, and I have a <a href="http://www.quartz-scheduler.org/docs/cookbook/JobInitPlugin.html" rel="noreferrer">jobs.xml file</a> which sets up my scheduled job that runs every minute.</p> <p>What I would like to do, is if the job is still running when the next trigger time rolls around, I don't want to start a new job, so I can let the old instance complete.</p> <p>Is there a way to specify this in jobs.xml (prevent concurrent instances)?</p> <p>If not, is there a way I can share access to an in-memory singleton within my application's <a href="http://www.quartz-scheduler.org/docs/api/org/quartz/Job.html" rel="noreferrer">Job</a> implementation (is this through the <a href="http://www.quartz-scheduler.org/docs/api/org/quartz/JobExecutionContext.html" rel="noreferrer">JobExecutionContext</a>?) so I can handle the concurrency myself? (and detect if a previous instance is running)</p> <hr> <p><strong>update:</strong> After floundering around in the docs, here's a couple of approaches I am considering, but either don't know how to get them to work, or there are problems.</p> <ol> <li><p>Use <a href="http://www.quartz-scheduler.org/docs/api/org/quartz/StatefulJob.html" rel="noreferrer">StatefulJob</a>. This prevents concurrent access... but I'm not sure what other side-effects would occur if I use it, also I want to avoid the following situation:</p> <p>Suppose trigger times would be every minute, i.e. trigger#0 = at time 0, trigger #1 = 60000msec, #2 = 120000, #3 = 180000, etc. and the trigger#0 at time 0 fires my job which takes 130000msec. With a plain Job, this would execute triggers #1 and #2 while job trigger #0 is still running. With a StatefulJob, this would execute triggers #1 and #2 in order, immediately after #0 finishes at 130000. I don't want that, I want #1 and #2 not to run and the next trigger that runs a job should take place at #3 (180000msec). So I still have to do something else with StatefulJob to get it to work the way I want, so I don't see much of an advantage to using it.</p></li> <li><p>Use a <a href="http://www.quartz-scheduler.org/docs/api/org/quartz/TriggerListener.html" rel="noreferrer">TriggerListener</a> to return true from vetoJobExecution().</p> <p>Although implementing the interface seems straightforward, I have to figure out how to setup one instance of a TriggerListener declaratively. <a href="https://stackoverflow.com/questions/2676514/quartz-documentation-for-xml-files">Can't find the docs for the xml file</a>.</p></li> <li><p>Use a <code>static</code> shared thread-safe object (e.g. a semaphore or whatever) owned by my class that implements Job.</p> <p>I don't like the idea of using singletons via the <code>static</code> keyword under Tomcat/Quartz, not sure if there are side effects. Also I really don't want them to be true singletons, just something that is associated with a particular job definition.</p></li> <li><p>Implement my own <a href="http://www.quartz-scheduler.org/docs/api/org/quartz/Trigger.html" rel="noreferrer">Trigger</a> which extends <a href="http://www.quartz-scheduler.org/docs/api/org/quartz/SimpleTrigger.html" rel="noreferrer">SimpleTrigger</a> and contains shared state that could run its own TriggerListener.</p> <p>Again, I don't know how to setup the XML file to use this trigger rather than the standard <code><trigger><simple>...</simple></trigger></code>.</p></li> </ol>

Querying!

Guidance

An individual column

Larger individual text columns get their own page to allow for proper reading.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload