Bug #22971
Updated by François ARMAND 11 months ago
In #12752 we added a test to check that when we save/delete a datasource, the datasource live instance is correctly reloaded (old killed, new spawned). On the CI, we sometime get: <pre> [2023-06-28T00:34:01.193Z] [info] operation from repository should [2023-06-28T00:34:01.829Z] [error] x saving rom repos should kill the old fiber [2023-06-28T00:34:01.829Z] [error] Running(RuntimeFlags(Interruption, CooperativeYielding, WorkStealing, FiberRoots), com.normation.plugins.datasources.HttpQueryDataSourceService.query(QueryService.scala:165)) is incorrect: (UpdateHttpDatasetTest.scala:909) [2023-06-28T00:34:01.829Z] [info] </pre> Which means that the fiber was not correctly killed and is still running. We need to know why it sometime happens: flacking test (timing issue?), or real problem ?