Closed jgsbennett closed 1 year ago
I think this is the checkpoint that is being processed at the point this all goes wrong: 20230711165044-875-10e97a79e2c10039375e23172427766b88640648b30b4673863e6a714179194b.zip
Can reproduce the failure by rerunning this job: https://jenkins.ops.vega.xyz/blue/organizations/jenkins/common%2Fsystem-tests-wrapper/detail/system-tests-wrapper/83024/pipeline/506 This runs exactly two tests in a very specific order, which reproduces the issue. (In case the job disappears from history, it is achieved by:
tests/governance/test_markets.py::test_settled_markets_saved_in_checkpoint_for_time_window
tests/governance/test_markets.py::test_propose_successor_on_expired_terminated_market
This also showed up in the snapshot soak run: https://jenkins.ops.vega.xyz/blue/organizations/jenkins/common%2Fsnapshot-soak-tests/detail/snapshot-soak-tests/8475/pipeline
The offending test is now skipped in: https://github.com/vegaprotocol/system-tests/pull/2224 So that we can get those new tests merged, so you'll need to turn it back on in a branch to see this fail.
Problem encountered
I can see a panic in the checkpoint restart code related to successor markets if I run all successor markets tests from jake_successor_markets test branch twice. (Guessing one of the later tests does something that the LNL test earlier in the run doesn't like, but I'll try to narrow that down).
Observed behaviour
Panic on second attempt
Expected behaviour
No panics. Two successful runs back to back
Steps to reproduce
Software version
03b02dc9fed3ba9d24a912426345ad179d45fc79
Failing test
No response
Jenkins run
No response
Configuration used
No response
Relevant log output