AEM as a Cloud Service Author outage due to MongoDB instability
AEM as a Cloud Service Author environments are temporarily unavailable when the underlying MongoDB database becomes unstable or shows high write latency. The environment recovers automatically after a short downtime. Monitor the Author environment and, if the issue recurs, collect the incident details and report them to Adobe Support.
Description description
AEM as a Cloud Service Author environments experience temporary outages if the underlying MongoDB database becomes unstable or exhibits high write latency. These conditions cause database contention, preventing the AEM Oak repository from updating its lease and leading to Author environment unavailability until recovery.
Environment
Adobe Experience Manager as a Cloud Service (Author environment)
Symptoms
- Temporary unavailability of the Author environment (Author rollout/pods become unavailable)
- Service recovers automatically after a short downtime (typically several minutes)
- Monitoring or logs may show MongoDB write latency spikes and high WiredTiger transaction-ticket utilization
Cause
The outage is caused by transient instability in the MongoDB primary/replica set and significantly elevated write latency. This leads to database contention, which prevents the AEM Oak repository from updating its lease, resulting in temporary termination of the Author process and unavailability until the environment recovers.
Resolution resolution
Follow these steps to monitor and report the issue.
-
Monitor the Author environment using available monitoring tools and logs for signs of instability, such as:
- Elevated MongoDB write latency
- High WiredTiger transaction-ticket utilization
- Author pods becoming unavailable or dropping to zero
-
If an outage occurs, confirm that the Author rollout recovers to a healthy state with all expected replicas available.
-
Continue monitoring the Author environment and MongoDB metrics for signs of recurrence.
-
If the issue recurs, collect the following information to assist with troubleshooting:
- Exact UTC timestamp of the incident
- Affected Author environment URL
- Observed error messages
- Scope of impact
- Relevant screenshots or request identifiers
-
Report the incident to Adobe Support with the collected information for further investigation.
To verify recovery, ensure that all Author pods are available and the environment is functioning as expected.
Related reading
AEM as a Cloud Service Author environment returns HTTP 503 and is inaccessible