[Questions] RabbitMQ queue crash after upgrading to 4.2 #15134
-
Community Support Policy
RabbitMQ version used4.2.1 Erlang version used27.3.x Operating system (distribution) usedWindows Server How is RabbitMQ deployed?Windows installer rabbitmq-diagnostics status outputNA Logs from node 1 (with sensitive values edited out)See https://www.rabbitmq.com/docs/logging to learn how to collect logs DetailsLogs from node 2 (if applicable, with sensitive values edited out)No response Logs from node 3 (if applicable, with sensitive values edited out)No response rabbitmq.confDefault Steps to deploy RabbitMQ clusterNA Steps to reproduce the behavior in questionNot reproducible advanced.configNA Application codeNo response Kubernetes deployment fileNo response What problem are you trying to solve?Could someone review the attached log file and advise if it indicates the root cause of the crash?
|
Beta Was this translation helpful? Give feedback.
Replies: 4 comments 2 replies
-
|
This is a known Windows-specific filesystem API behavior that we have addressed in some places but possibly not all. A file handle on Windows cannot be closed if it has more than one concurrent user or something like that. This time the function that has run into it is Switching to running RabbitMQ on Linux, which is something our team strongly recommends, will fundamentally address this behavior. |
Beta Was this translation helpful? Give feedback.
-
|
Searching this repository's issues, discussions and PRs for |
Beta Was this translation helpful? Give feedback.
-
|
The problem is that after the crash the queue restarts and tries to recover fast enough that opening this same file gives the same error as the original crash. And then it gives up. We can add an |
Beta Was this translation helpful? Give feedback.
-
|
I'm adding another crash log from today to help further identify the error. Details |
Beta Was this translation helpful? Give feedback.
I have opened #15136 with this change.