Closed dvs001-ifs closed 4 months ago
I'm not sure that everyone would agree that this is not an error. If you tag an instance with a schedule, you indicate your intention for it to be started/stopped according to that schedule. If instance scheduler cannot do that, that is a failure in the operation of the solution - though not a fatal one, since it can continue to schedule other resources. It might be surprising to other customers that instance scheduler would not consider that an error. Is there a reason you are unable to a) remove the tag from the instances you don't wish to schedule, or b) tag multi-az RDS instances with a schedule with a 24-hour running period (effectively a "no-action" schedule)?
However, we would like to explore improved reporting of failure conditions, and I believe that the error message for this specific issue is improved in v1.5.0:
Error stopping rds instance RDS:database-1-instance-1, (An error occurred (InvalidParameterCombination) when calling the StopDBInstance operation: You can’t stop a DB instance that is part of a Multi-AZ DB cluster.)
Closing the issue since we decided to not implement. Please reach out again if you have more questions thanks.
Describe the bug ERROR message is logged when the scheduler is unable to about stop multi-az RDS instances. If this is a limitation, then can scheduler log the message as a WARNING instead of an ERROR message?
Use case for keeping it in the schedule:
If the RDS is multi AZ, scheduler reports ERROR in logs and our alarms go off. Error I receive: Error stopping rds instance RDS:, (Shape is missing required key 'type': OrderedDict([('members', OrderedDict([('SourceRegion', OrderedDict([('shape', 'String'), ('documentation', '
The ID of the region that contains the source for the db instance.
')]))]))]))Would be nice if inability to shutdown multi-AZ RDS instance is logged as a WARNING, since this is a known limitation of scheduler.
Version: 1.4.2