apache / apache/incubator-seata
GlobalSession has concurrency problem in cluster mode?
Open
type: question
- Dominant language
- Java
- Stars
- 26k
- Forks
- 8.8k
- Avg merge
- 1d 8h
- Merged PRs (30d)
- 4
Description
## Why you need it?
Is it possible that more than one servers currently changing globalSession's status in cluster mode? If this situation happened, local ReentrantLock can not keep globalSession status changing procedure thread safe. Is it correct?
e.g., server A is commiting a transaction, meanwhile server B's timeout checking thread is trying to start a rollback. Is this a problem?
## How it could be?
Is optimistic locking on global_table a possible solution?
Contributor guide
Assessment
This issue has not been assessed yet.