-
Bug
-
Resolution: Done
-
Normal
-
OpenShift 4.7 Async
-
None
-
devex docs #207 Sep 9-Sep 30
-
3
-
Documentation (Ref Guide, User Guide, etc.)
-
Undefined
-
---
-
---
[ISSUE]
In the operations is indicated to do this step:
3. Perform a shard synced flush using the {product-title} es_util tool to ensure there are no pending operations waiting to be written to disk prior to shutting down:
But, if we don't stop to Elasticsearch for receiving data from fluentd or even from Kibana, then, at the same time that we are executing that command new write operations to the disk can be waiting, then, I'll add an extra step before (confirm this with engineering) that it should be stop the fluentd pods of sending logs to Elasticsearch, for doing it, we could add before the current 3 step, this:
3. Scale down the fluentd pods to avoid them to send new logs to the Elasticsearch $ oc -n openshift-logging patch daemonset/logging-fluentd -p '{"spec":{"template":{"spec":{"nodeSelector":{"logging-infra-fluentd": "false"}}}}}'
And the latest step of all of of the commands after the current 7:
8. Scale down the fluentd pods to avoid them to send new logs to the Elasticsearch $ oc -n openshift-logging patch daemonset/logging-fluentd -p '{"spec":{"template":{"spec":{"nodeSelector":{"logging-infra-fluentd": "true"}}}}}'
