What could cause Coordinator instance to stop being a Leader?

Hi, All.

I am currently face a puzzling situation where in my current Druid configuration only one instance of Coordinator is allocated.
At one point, it is temporarily stop being the Leader. I wonder what could be the possible cause of that?

I am currently running Druid 0.16.0.

Hi Keyne,
Can you give more details about your cluster(how many master nodes, query nodes ,data nodes)?

Any clues(errors/warnings/failures/exceptions) from your coordinator logs?

Thank you,

–siva

Hi Siva,

Currently, this is the configuration for the Druid cluster:

  • Coordinator: 1 instance
  • Overlord: 1 instance
  • Broker: 1 instance
  • Historical: 4 instances

They are running on ECS and so far I only get the docker log which did not goes as far to the point when the issue arises.
I failed to locate the actual logs within the running docker container itself, maybe if you can show some light on which folder I should look or which configuration I should update?

Thank you very much in advance.

Load balancer

U.S. BANCORP made the following annotations---------------------------------------------------------------------

Hi keene,
Generally logs are created in var/sv// under druid home directory . For example, coordinator log should be at var/sv/coordinator/ under druid home directory.

or

Try to run find command to search for filename “current” and see whether you will be able to locate log files created by druid.

Thanks,

–siva