ack-sel is not very granular
Bug #1928403 reported by
Andrea Ieri
This bug affects 4 people
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
hw-health-charm |
Fix Released
|
Wishlist
|
Mert Kirpici |
Bug Description
In bug 1901735 we implemented the ack-sel action by relying on the --date-range ipmi-sel option. This option does not accept anything more granular than whole days, and as a consequence it is not possible to silence an alert produced by a SEL entry that is less than a day old.
It would be useful to find a workaround, as operators are currently forced to downtime alerts for at least one day before they are able to silence them via this action.
Related branches
~mertkirpici/charm-hw-health:lp/1928403
- Eric Chen: Approve
- Facundo Ciccioli: Approve
- 🤖 prod-jenkaas-bootstack (community): Approve (continuous-integration)
- JamesLin: Approve
- BootStack Reviewers: Pending requested
-
Diff: 483 lines (+153/-136)6 files modifiedsrc/actions/actions.py (+31/-13)
src/lib/hwhealth/tools.py (+16/-10)
src/tests/functional/test_hwhealth.py (+0/-98)
src/tests/unit/test_actions.py (+92/-0)
src/tests/unit/test_ipmi_sensor.py (+14/-14)
src/tox.ini (+0/-1)
Changed in charm-hw-health: | |
importance: | Undecided → Wishlist |
tags: | added: bseng-469 |
Changed in charm-hw-health: | |
status: | New → In Progress |
assignee: | nobody → Mert Kirpici (mertkirpici) |
Changed in charm-hw-health: | |
status: | In Progress → Fix Committed |
Changed in charm-hw-health: | |
milestone: | none → 23.01 |
Changed in charm-hw-health: | |
status: | Fix Committed → Fix Released |
To post a comment you must log in.
If today is Jan 05 2022, ack-sel will render this in cron job:
--seloptions --date- range=01/ 05/2022- now
However, this will still include the events occurred on 2022-01-05, and fail to mute current alert.
A few ways to workaround:
1) use 01/06/2022 as start date. date-range= 01/01/1970- 01/05/2022` instead display- range=0- 42` instead.
2) use `--exclude-
3) get the last event id, e.g.: 42, and use `--exclude-
for 1) and 2), it will have side effect to also ignore following alerts in the same day.
3) seems to be accurate, but, it looks like, when we clear sel, the id will go back to 0. So new generated alerts 0-42 may be ignored unexpected. Need to confirm this.