[IOTDB-4222] Create data region repeatedly when concurrently deleting and creating storage groups

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.13.0, 0.14.0
• Fix Version/s: 0.14.0, 0.13.2
• Component/s: Core/Engine
• Labels: pull-request-available
• Sprint: 2022-8-Cluster
• Assignee: Haiming Zhu
• Reporter: Haiming Zhu
• Created: 24/Aug/22 09:55
• Updated: 25/Aug/22 03:16
• Resolved: 25/Aug/22 03:16

Description

One data region maybe created repeatedly when the storage group is created and deleted concurrently.

The procedure is (threads are conccurent):

1. thread-1 creates one storage group
2. thread-2 writes some data
3. thread-3 deletes this storage group
4. thread-4 creates this storage group again

The root cause is as below:

When the first thread creates one storage group and the data region hasn't been ready, the second thread deletes this strorage group and the third thread creates this storage group again. The two create procedures's mnodes are different, so it's unsafe to synchronize mnode.

Issue Links

• links to: GitHub Pull Request #7111
• links to: GitHub Pull Request #7112