2cb728678f
The mtcAgent is seen to get stuck handling a subfunction failure
detected over self (controller-0) unlock of an AIO simplex controller.
It gets stuck reporting that is it already handling the failure,
but isn't.
log flooding: 'controller-0 already handling force full enable'
This issue only exists in AIO simplex when the subfunction enable
handler detects the failure. This issue was introduced by the
following update:
Remove Start Host Service Launch in mtcAgent & enhance fault detection
https://opendev.org/starlingx/metal/commit/
6106051f1c
Test Plan:
PASS: Verify an AIO simplex self unlock subfunction failure leads to
'degrade' state with 'enable failure' alarm.
PASS: Verify same issue for the standby controller leads to
'failure' state with 'enable failure' alarm.
Regression:
PASS: Verify spontaneous unhealthy active controller is degraded.
PASS: Verify spontaneous unhealthy standby controller is failed.
Closes-Bug: 2119449
Change-Id: I5ab5e6d85906f1923a0828211dbf94d2f82e73f8
Signed-off-by: Eric MacDonald <eric.macdonald@windriver.com>