apache / apache/incubator-pegasus
Fix time-consuming CI tests
- Dominant language
- C++
- Stars
- 2.1k
- Forks
- 328
- PR merge metrics
- No merged PRs in 30d
Description
Here I list the most time-consuming unit tests in rDSN:
- [ ] meta_load_balance_test.simple_lb_cure_test (14650 ms)
- [ ] meta.data_definition (40765 ms)
- [ ] meta.adjust_dropped_size (11367 ms)
- [ ] meta.state_sync (39060 ms)
- [ ] meta.update_configuration (14376 ms)
- [ ] meta.balancer_validator (60094 ms)
- [ ] fd.worker_died_when_switch_master (12001 ms)
- [ ] fd.switch_new_master_suddenly (18002 ms)
- [ ] fd.dummy_connect_disconnect (11002 ms)
- [x] tools_common.asio_network_provider_connection_threshold (70017 ms)
All of them spend in total 5min. It's unacceptable to have these tests taking so long. Welcome if anyone could help me to fix any of them.
Contributor guide
Research direction
The issue identifies slow tests including meta_load_balance_test.simple_lb_cure_test, meta.data_definition, meta.state_sync, meta.balancer_validator, and several fd tests, but no source files. Start by locating and running one listed test to determine whether the delay is caused by setup, waits, or the test body. Done means the selected test is faster without losing coverage and the relevant CI tests still pass.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- cpp
- Domain
- performance, testing
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100