Bug #5727
closedNodes stays in unexpected/missing status
Description
In compliance mode, my nodes are almost always in missing/unexpected status.
It seems that the reports are not linked to a config id:
here is an extract of what i got from my dev environment (file compliance_reports)
I really don't think it's related to clock (my reports are consistent with my dev webapp)
Files
Updated by Jonathan CLARKE over 9 years ago
Is this still broken? If so, it needs fixing ASAP!
Updated by Jonathan CLARKE over 9 years ago
- Target version changed from 3.0.0~beta1 to 3.0.0~beta2
Updated by François ARMAND over 9 years ago
Starting with a fresh set of data, we weren't able to reproduce it like that.
None the less, we were able to see intermitent phase were the reports were switching from a normal displaying (mostly success/no answer/errors) to a only missing/unexpected one.
The interpretation from code point of view is that at some point, we must be expecting reports for a given version of the policy configuration and got reports for an other one (so that all expected reports were missing and we get a bunch of unexpected ones). Given the transitionnal and periodic charachterisation of the state, I would guess that some computation regarding the lenght of a era where a give configuration is OK are buggy. For now, I didn't successed in crafting failing test nor reproducing the pb in a consistent manner.
Updated by François ARMAND over 9 years ago
One other strange behaviour with unexpected/missing reports:
A rule R1 exists, with D1 / G1.
A rule R2 is created, also with D1 / G1.
Then, R2 is correclty displayed as "applying", but R1 display "missing reports"
Updated by François ARMAND over 9 years ago
- Target version changed from 3.0.0~beta2 to 3.0.0~rc1
Updated by Nicolas CHARLES over 9 years ago
François ARMAND wrote:
One other strange behaviour with unexpected/missing reports:
A rule R1 exists, with D1 / G1.
A rule R2 is created, also with D1 / G1.
Then, R2 is correclty displayed as "applying", but R1 display "missing reports"
this looks a lot like http://www.rudder-project.org/redmine/issues/6020
Updated by Nicolas CHARLES over 9 years ago
This may be fixed by https://github.com/Normation/rudder/pull/741 , but since we couldn't reliably reproduce it, i am not sure
Updated by François ARMAND about 9 years ago
It seems more likely that the problem was the missing "complete == true" when looking for new run information, corrected in https://github.com/Normation/rudder/pull/743 (#6040)
Updated by Vincent MEMBRÉ about 9 years ago
- Target version changed from 3.0.0~rc1 to 3.0.0
Updated by François ARMAND about 9 years ago
- Status changed from New to Rejected
That wasn't observe anymore since the correction of #6040, so closing it as resolve (duplicate)