Eric Pai created IOTDB-4210:
-------------------------------
Summary: Dead lock during WALNode recover
Key: IOTDB-4210
URL: https://issues.apache.org/jira/browse/IOTDB-4210
Project: Apache IoTDB
Issue Type: Bug
Components: Core/WAL
Reporter: Eric Pai
Assignee: Eric Pai
Fix For: master branch
Attachments: image-2022-08-23-11-21-47-042.png,
image-2022-08-23-11-23-39-791.png, image-2022-08-23-11-24-00-430.png
In our environment, we met a dead lock bug during start up. And here is the
stacktrace of two suspect threads:
* Thread 1: Normal client PRC thread
!image-2022-08-23-11-24-00-430.png!
* Thread 2: WAL recover thread (maybe the root cause)
!image-2022-08-23-11-21-47-042.png!
We can see that the thread1 is blocked by wanting to enter the sychronized
block of IStorageGroupNode, and thread2 has already holded the same object, but
waiting for the WALRecoverListener to be notified. However, there's no other
thread to try to notify it.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
