掲示板
Scheduler Trigger Issue
8年前 に Sai Mohan Gadamsetty によって更新されました。
Scheduler Trigger Issue
Junior Member 投稿: 92 参加年月日: 13/02/06 最新の投稿
Hi All,
We have Liferay 6.1.2 on Prod, setup with a cluster (2 nodes). We need to run a scheduler every day 00:30 hrs (GMT) to send emails to certain users with an update. So, I configured trigger time in scheduler as below,
<scheduler-entry>
<scheduler-event-listener-class>com.liferay.scheduler.WhatsNewScheduler</scheduler-event-listener-class>
<trigger>
<cron>
<cron-trigger-value>0 30 0 * * ?</cron-trigger-value>
</cron>
</trigger>
</scheduler-entry>
This configuration working exactly fine and delivering emails to end users as expected. But today this scheduler has run 2 times 15:35 (GMT) and following normal scheduler time 00:30. Why Liferay ran scheduler 2 times. I found some logs before scheduler started first time,
can any one help me why Liferay triggered like this?
We have Liferay 6.1.2 on Prod, setup with a cluster (2 nodes). We need to run a scheduler every day 00:30 hrs (GMT) to send emails to certain users with an update. So, I configured trigger time in scheduler as below,
<scheduler-entry>
<scheduler-event-listener-class>com.liferay.scheduler.WhatsNewScheduler</scheduler-event-listener-class>
<trigger>
<cron>
<cron-trigger-value>0 30 0 * * ?</cron-trigger-value>
</cron>
</trigger>
</scheduler-entry>
This configuration working exactly fine and delivering emails to end users as expected. But today this scheduler has run 2 times 15:35 (GMT) and following normal scheduler time 00:30. Why Liferay ran scheduler 2 times. I found some logs before scheduler started first time,
can any one help me why Liferay triggered like this?
添付ファイル:
8年前 に Andrew Jardine によって更新されました。
RE: Scheduler Trigger Issue
Liferay Legend 投稿: 2416 参加年月日: 10/12/22 最新の投稿
Hi Saj,
I once had this problem on 6.1 as well (GA2 I believe it was). In my case i was using the API to create job dynamically but found that the jobs we're running incrementally (even worse than your case). So if I had 5 jobs scheduled to run at 00:30 .. then the first one would run 5 times, the second 4 times, the third 3 times etc.
Unfortunately I never found the root cause of the issue for me and in the end the client abandoned the feature so I stopped investigating. I don't have an actual solution to your problem (I'm sorry) but I wanted to let you know that someone else experienced issues with 6.1 and jobs (only in a cluster though). That was the oddest part, on my local development environment, it was never an issue.
I do recall that there are several tables (all prefixed with QUARTZ_) in the database. Have you had a look into those tables to see if there are any odd entries? The only other thing I can think of is -- have you deployed your portlet a second time? perhaps rather than replacing the initial job, it has appended another? -- though that would certainly not explain the time difference from when it was run. That could reall only be explained by a timezone setting.
I'm happy to try to help you with this issue (debugging it) -- are you able to share your source?
I once had this problem on 6.1 as well (GA2 I believe it was). In my case i was using the API to create job dynamically but found that the jobs we're running incrementally (even worse than your case). So if I had 5 jobs scheduled to run at 00:30 .. then the first one would run 5 times, the second 4 times, the third 3 times etc.
Unfortunately I never found the root cause of the issue for me and in the end the client abandoned the feature so I stopped investigating. I don't have an actual solution to your problem (I'm sorry) but I wanted to let you know that someone else experienced issues with 6.1 and jobs (only in a cluster though). That was the oddest part, on my local development environment, it was never an issue.
I do recall that there are several tables (all prefixed with QUARTZ_) in the database. Have you had a look into those tables to see if there are any odd entries? The only other thing I can think of is -- have you deployed your portlet a second time? perhaps rather than replacing the initial job, it has appended another? -- though that would certainly not explain the time difference from when it was run. That could reall only be explained by a timezone setting.
I'm happy to try to help you with this issue (debugging it) -- are you able to share your source?