Bolt Task Scheduler running at 11am instead of 3am - cron

I've been monitoring the cron jobs I set up last week from my Bolt powered website. I've noticed that the daily cron jobs seem to be running at 11am (that's the time the database logs them at). In the config.yml the time is set to 3am.
I've checked the server time and that's using UTC. The MySQL database is using the server time so I would assume that to be UTC as well.

Is this on a Git master version, or 1.x?
Either way this definitely sounds like a bug.
Edit:
Was a bug and I have submitted a PR that fixes it for master and the version 1.6 branch.
However a possible workaround is to set the time of the existing records in the database so that the time of the last run is 03:00.

Related

Issue after Quartz Web Restart- Quartz triggers all the scheduled jobs which has been pasted

I setup JDBCJobStore for store jobs and schedule the jobs by Cron.
Sometimes, I will manually stop the Quartz Scheduler in order to bypass some scheduled job to be triggered for some specific purpose.
However, I face an issue after re-starting the Quartz Scheduler. All the jobs which was scheduled will be triggered at same time even through the next schedule time has been due. I check the database and find all the jobs has been scheduled and saved in QRTZ_FIRED_TRIGGERS table, but not can be delete. Cron only re-schedule jobs after run.
Is there any way to make Quartz to re-schedule job by Cron when I re-start the Quartz Server and without trigger these expired schedule?
Any help will be highly appreciated.
Best Regards,
Dean Huang
The job will be reschedule by Cron if I configure as RAMJobStore and setup job by xml, but not JDBCJobStore.

Restart Apache Zeppelin Interpreter from Paragraph

I want to restart spark interpreter in zeppelin from a paragraph. Currently, due to the timeout, the interpreter gets stopped automatically after 100 minutes of inactivity. However, as I want a scheduled execution of some notebooks once a day I can't use a GUI solution (restart interpreter from config manually).
I know that one solution would be to increase the timeout, but I'd rather restart it on demand than have it on all the time.

Why can't I cancel CouchDB 2.1.1 replication?

Just a couple days ago, I migrated from CouchDB 1.6.1 to 2.1.1 in our production environment. The couchup replicate step was taking forever, and then I noticed that most continuous pull replication was running, so I figured that must be slowing it down.
So I deleted all the continuous replication documents from the _replicator database, hoping that would stop the replication. It didn't. I even restarted CouchDB, and the continuous replication is still occurring.
If I look in the Replication section in Fauxton, both tabs are empty. But when I look in the Tasks section, continuous replication is still running, 3 days later.
How do I stop continuous replication?
===========================================
UPDATED 2018/03/27 21:00 UTC
Trying to delete from _scheduler, but that's not allowing me to do anything but GET and HEAD. And I'm logged in as the system admin. (FWIW, _scheduler is not showing up as a database).
# curl -k -X GET 'http://localhost:5984/_scheduler/jobs/5e9bf92325eee5caff026e47197f9c3d+continuous+create_target' -u couchdbadmin:XXXXXXXX
{"database":"_replicator","id":"5e9bf92325eee5caff026e47197f9c3d+continuous+create_target","pid":"<0.606.0>","source":"http://couchdbadmin:*****#prdpccggww3n03.w3-969.ibm.com:5984/youribm-timeinout/","target":"http://couchdbadmin:*****#localhost:5984/youribm-timeinout/","user":null,"doc_id":"youribm-timeinout-continuous-replication-from-primary","history":[{"timestamp":"2018-03-24T00:01:41Z","type":"started"},{"timestamp":"2018-03-24T00:01:41Z","type":"added"}],"node":"couchdb#127.0.0.1","start_time":"2018-03-24T00:01:41Z"}
# curl -k -X DELETE 'http://localhost:5984/_scheduler/jobs/5e9bf92325eee5caff026e47197f9c3d+continuous+create_target' -u couchdbadmin:XXXXXXXX
{"error":"method_not_allowed","reason":"Only GET,HEAD allowed"}

CPU/Mem usage for cron jobs on GCP K8S while not running

I have setup a cron job on GCP Kubernetes. It runs once per day at 10:00am.
The job runs as expected however, I don't really understand what the charts say on the GCP K8S console.
As shown in the charts, there are around 1.5 CPU and 8G RAM at this point, when the cron job is not running. I expect the current usage should be zero as it is not running.
Could anyone see what is wrong? or I read the charts wrong?
Note I do retain 7 jobs in the history. Each job ran about 15sec and completed successfully.
State: Terminated
Reason: Completed
Exit Code: 0
Started: Thu, 08 Mar 2018 04:00:56 +0000
Finished: Thu, 08 Mar 2018 04:01:09 +0000
20180320 EDIT:
I found all graphs from other cron jobs all look the same. Is it something I setup wrong?
I reproduced your situation by that way:
Created CronJob with a name "Application"
Added an application using Deployment with a name "Application"
And now, I see on graphics of Cron Job details information not only about a cronjob itself, but also about an application.
So, I think, dashboard using name of object for getting a data.
Looks like you have any other deployment, replicaset etc. which always running and has a same name as your cronjob and on graphics you see the mix of data from it and from your cronjob.

How can I set the time-zone for the Spark History Server?

When I submit a job on a set of machines machine located in PST timezone, the Spark Master has the correct time, but the history server shows time that is 8 hrs ahead which is GMT.
Is there a way to fix this ?

Resources