Recurrent insert errors in database on vm schedules #11564
Description
problem
After creating a schedule to start/stop vm_instances, a recurrent error appears in logs
025-07-17 11:29:19,383 ERROR [o.a.c.v.s.d.VMScheduledJobDaoImpl] (VMSchedulerPollTask:[ctx-d11d1302]) (logid:c3157b08) DB Exception on: HikariProxyPreparedStatement@614529765 wrapping com.mysql.cj.jdbc.ServerPreparedStatement[6379]: INSERT INTO vm_scheduled_job (vm_scheduled_job.uuid, vm_scheduled_job.vm_id, vm_scheduled_job.vm_schedule_id, vm_scheduled_job.async_job_id, vm_scheduled_job.action, vm_scheduled_job.scheduled_timestamp) VALUES (x'65396330306236652d626463622d346430612d396162332d663466636136343536356130', 293, 120, NULL, 'STOP', '2025-07-18 04:00:00') java.sql.SQLIntegrityConstraintViolationException: Duplicate entry '120-2025-07-18 04:00:00' for key 'vm_schedule_id'
I have two management host running with ACS 4.20.1 with KVM Hypervisors. Any clue what is trying to insert it again in the database?
versions
Version
ACS 4.20.1
KVM Hyps
Ceph Storage
VPC Tiers
The steps to reproduce the bug
- Have 2 ACS management server running
- Create a VM
- Schedule START / STOP
What to do about it?
It appears to be a race condition error caused by multiple scheduler instances running concurrently. We have two ACS Server Manager nodes, and when a scheduled task is created on one node, the other node throws an error — and the same happens in reverse.