Dx NetOps Unified Collection Framework Error: "cannot execute SELECT FOR UPDATE in a read-only transaction" after Postgres failover
Article ID: 441322
Updated On:
Products
Issue/Introduction
Postgres failover triggers read-only transaction errors, causing the [scheduler-pod] to remain stuck and halting polling across all plugins․
ERROR MESSAGES:
"MisfireHandler: Error handling misfires: Failure obtaining db row lock: ERROR: cannot execute SELECT FOR UPDATE in a read-only transaction"
"ClusterManager: Error managing cluster: Failure updating scheduler state when checking-in: ERROR: cannot execute UPDATE in a read-only transaction"
Environment
UCF Version 25.4.8 and earlier
Cause
A Postgres role switch leaves the scheduler pod in a stuck state unable to obtain database row locks․
Resolution
RESTART THE AFFECTED POD
Command: kubectl delete pod [scheduler-pod] -n [namespace]
EXPECTED: The pod restarts and polling resumes
NOTE: Replace [scheduler-pod] with the actual pod name
VERIFY SUCCESS:
Polling resumes for plugins
No read-only transaction errors appear in the logs
Additional Information
This condition will be resolved with UCF v25.4.9+
Feedback
