Summary
On July 14, 2026, between 14:40 UTC and 17:00 UTC, some customers experienced AI Agents getting stuck in loops and sending repeated messages in bot conversations. In some cases, this continued even after an end user stopped responding or after a human agent replied.
Timeline
We are aware of an issue affecting bot conversations causing looped or repeating messages across multiple pods. Our team is investigating and we will provide further updates shortly.
We have rolled back a recent change which appears to have caused messaging conversations to loop or repeat bot messages, and we are seeing conversation errors drop off. Please let us know if you are still experiencing any issues in conversations.
We are happy to report that the issue causing repeated and looping messages in bot conversations has been resolved. Thank you for your patience during our investigation.
Root Cause Analysis
A change to a cleanup job increased query timeouts to accommodate a growing backlog. Under real production load, the query ran long enough to hold locks on key database tables, blocking the live webhook processing path.
This lock contention resulted in database “lock wait timeout” errors during inbound webhook processing, preventing conversation state from being updated and leading to repeated/looping bot behavior.
Resolution
- We confirmed the issue via logs and reproduction in a test account.
- We rolled back the cleanup job change.
- After rollback, conversation processing errors dropped and customer reports indicated the looping/repeating behavior stopped.
Remediation Items
We are taking steps to reduce the likelihood of recurrence and improve earlier detection:
- Query and job improvements: Rework and optimize the cleanup process so it can handle backlog safely without locking behavior that impacts live traffic.
- Monitoring and alerting: Add/adjust monitoring so we detect job failures, rising lock-wait timeouts, and related error patterns earlier.
-
Operational safeguards: Review rollout safeguards for changes that can affect database performance or locking, especially for background jobs sharing the same database as live request paths.
FOR MORE INFORMATION
For current system status information about Zendesk and specific impacts to your account, visit our system status page. If you have additional questions about this incident, contact Zendesk customer support.