URL: https://github.com/SSSD/sssd/pull/947 Title: #947: tests: fix race conditions in integration tests
alexey-tikhonov commented: """
Sorry if I was not clear enough. I was talking about different bug/race: when updated data are requested _before_ files_provider even started update. 1. edit actual files (/etc/passwd...) 2. files_provider notices the change (via inotify) and starts internal update 3. updated is finished, provider is ready
You are talking about request during (2). I am talking about request between (1) and (2). I believe this is a reason of additional sleeps and hides a bug.
And yes, mem-cache actually adds complexity to this issue. """
See the full comment at https://github.com/SSSD/sssd/pull/947#issuecomment-559463387