I have two Jobs (for Bix extract) configured to run 1 time at an specific hour, but is running twice
I have deployed two new Jobs configured to run from Sunday to Friday at 10:45 pm EST (approximately). Those Jobs run an activity to extract some bix files. Since the EST time change, the Jobs have been running twice a day with one hour of difference. Is happening in production and lower environments.
I don't know why this is happening and how to fix it. Could you help me?
Bits in Glass
IN
@LauraC59 Can you refer this below URL which says about the same issue? https://support.pega.com/question/scheduler-configured-run-any-one-associated-node-running-all-nodes?
Till 8.4--> Run on any one associated node was running the job scheduler on only 1 node. But in 8.6 with same configuration job scheduler is running on all available nodes. It is due to split brain mechanism is what post conveys. They suggest us to retart all the nodes so the nodes will start to communicate each other after restarting and only 1 node will run the job scheduler and this will be communicated to other nodes.
Scotiabank
CO
Hi @Anoop Krishna.
I did the restart on the low environment. It run at 11:45 AM, but the next run said 12:45 PM, which means that will run twice again.
Do you think that maybe the issue is for the time zone? in the jobs, I use Canada/Eastern, but the Pega app is configured in UTC
Bits in Glass
IN
@LauraC59that also can be an issue. Maybe we can check these 2 scenarios:-
1- with same timezone 2. In the job scheduler activity, you can set node id and send notification or log message so that we can verify whether same node is running twice or not.
Scotiabank
CO
@Anoop Krishnais a bix extract, so we can see in the extract that was run by two different nodes with a difference of one hour
Bits in Glass
IN
@LauraC59 ok so we identified that it is run by 2 different nodes. But job scheduler configuration is "run on any one associated node" right?
Scotiabank
CO
@Anoop Krishna Right
Bits in Glass
IN
@LauraC59 ok then that's the issue. Your nodes are not communicating to each. I have posted a link above which says about split brain issue, like 1 node will not be identify if other node exists or active. If your restart didn't work you will have to raise SR. Complete restart should have worked, if not would need to go with SR.