Details
-
Bug
-
Resolution: Done
-
Minor
-
Logging 5.4.z
-
False
-
None
-
False
-
NEW
-
OBSDA-7 - Adopting Loki as an alternative to Elasticsearch to support more lightweight, easier to manage/operate storage scenarios
-
VERIFIED
-
Before this update, <X problem> caused <Y situation> (OPTIONAL: under the following <Z conditions>). With this update, <fix> resolves the issue (OPTIONAL: and <agent> can <perform operation> successfully).
-
Log Storage - Sprint 221
Description
Customer needs to be aware that logs cannot be sent to Loki through 'http' after recent changes in how the route is handled.
This is for Loki upgrade from 5.4.2 to 5.4.3
Steps to reproduce:
1) Install Loki Operator v5.4.2 from redhat-operators catalog source. Create storage bucket and LokiStack CR.
[kbharti@cube hack]$ oc get csv -n openshift-logging
NAME DISPLAY VERSION REPLACES PHASE
cluster-logging.5.4.3 Red Hat OpenShift Logging 5.4.3 Succeeded
loki-operator.5.4.2-30 Loki Operator 5.4.2-30 Succeeded
2) Forward logs to Loki using bearer token and CLF.
3) Logs can be queried through logcli
logcli -o raw --bearer-token="$(oc whoami -t)" --addr "http://lokistack-dev-openshift-logging.apps.kbharti-411-142.qe.devcluster.openshift.com/api/logs/v1/application" query '{kubernetes_namespace_name="test"}'
4) Upgrade Loki from 5.4.2 to 5.4.3 using new cpaas index image and catalog source.
Index image: quay.io/openshift-qe-optional-operators/ocp4-index:cpaas-logging-1653
5) New logs are not present on Loki
To send logs to Loki after 5.4.3 upgrade, follow below guide
https://github.com/grafana/loki/blob/main/operator/docs/forwarding_logs_to_gateway.md