[RATIS-2151] TestRaftWithGrpc keeps failing with zero-copy

• Type: Sub-task
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 3.2.0
• Component/s: gRPC, server
• Assignee: Tsz-wo Sze
• Reporter: Tsz-wo Sze
• Created: 07/Sep/24 02:13
• Updated: 11/Oct/24 16:59
• Resolved: 11/Oct/24 15:01

Description

TestRaftWithGrpc keeps failing with zero-copy. It seems that RATIS-2129 makes it much easier to fail. Also, probably due to bugs is LeakDetector, some failures may not be detected without RATIS-2164.

• Master branch after RATIS-2129: TestRaftWithGrpc#[781d61d37411b374f104eb0806e1e2c4090fb35e]-10x10: 91/100 failures
  https://github.com/szetszwo/ratis/actions/runs/10747241634/job/29810232738

• Master branch before RATIS-2129: TestRaftWithGrpc#[dfed1012983d1d7b5fb2c408e19b8661cbe000b4]-10x10 success
  https://github.com/szetszwo/ratis/actions/runs/10746526581

Note that it won't fail when applying RATIS-2129 to 3.1.1:

• 3.1.1 after RATIS-2129:TestRaftWithGrpc#[76003e4113d98ff09663d384a020ce7e39b5b323]-10x10 success
  https://github.com/szetszwo/ratis/actions/runs/10798194952

Issue Links

• is duplicated by: RATIS-2159 TestRaftWithSimulatedRpc could "fail to retain".
• relates to: RATIS-2129 Low replication performance because of lock contention on RaftLog
• links to: GitHub Pull Request #1164