A DevOps engineer notices that a system occasionally gets slower.
• No deployment has been recently performed.
• After a restart of the instance, the performance degradation is mitigated.
• If no action is taken, the JVM tends to crash in an unpredictable manner with Out Of Memory exceptions.
• There is no clear correlation between visitor peaks and crashes of the instances.
• The same symptoms affect author and publish instances.
What is causing this issue?
Submit