Before sending notification verifying no changes in alert for sometime #178
Conversation
GowthamShanmugam
changed the title
Codecov Report
@@ Coverage Diff @@
## master #178 +/- ##
==========================================
+ Coverage 23.26% 23.64% +0.37%
==========================================
Files 12 12
Lines 417 423 +6
Branches 50 52 +2
==========================================
+ Hits 97 100 +3
- Misses 315 317 +2
- Partials 5 6 +1
Continue to review full report at Codecov.
|
tendrl-bug-id: Tendrl#177 bugzilla: 1564175 Signed-off-by: GowthamShanmugasundaram <gshanmug@redhat.com>
GowthamShanmugam
force-pushed
the
fix_notifications
branch
from
401bb45
to
b817ed6
Compare
| lock.release() | ||
| if (datetime.utcnow() - parser.parse( | ||
| alert.time_stamp | ||
| ).replace(tzinfo=None)).seconds >= 120: |
There was a problem hiding this comment.
it won't sent notification if any changes in alert for last 120 sec. And i have verified "dateutil" package is there in centos and rhel and fedora also
There was a problem hiding this comment.
How we decide 120 sec is enough and foolproof to wait for?
GowthamShanmugam
changed the title
|
@r0h4n @shtripat @nthomas-redhat please review |
|
@fbalak if possible please verify and comment here |
…t status for sometime tendrl-bug-id: Tendrl#177 bugzilla: 1564175 Signed-off-by: GowthamShanmugasundaram <gshanmug@redhat.com>
…t status for sometime tendrl-bug-id: Tendrl#177 bugzilla: 1564175
|
@shtripat grafana sometimes changing alert status in 60 sec but raising notification after 120 sec, it is not happening always but sometimes it is happening. Warning dashbaord raise alert raised info alert after 60 sec when threshold cross 90 and above, but critical raise alert after 120 sec only. i have checked the alert history they modified alert status in 60 sec but notification came after 120 sec only. |
tendrl-bug-id: #177
bugzilla: 1564175
Signed-off-by: GowthamShanmugasundaram gshanmug@redhat.com