-
Bug
-
Resolution: Done
-
Undefined
-
None
-
4.13
-
No
-
CNF RAN Sprint 233, CNF RAN Sprint 234, CNF RAN Sprint 235, CNF RAN Sprint 236
-
4
-
False
-
-
-
Rel Note for Telco: Not Required (4.12) - this has moved to 4.13 (see note from Yang below)
This is a clone of issue OCPBUGS-2832. The following is the description of the original issue:
—
Starting with a working BMER setup, using http sidecar and a consumer, the consumer gets the events.
I then restart the hw-event-proxy pod, and wait until the new pod is ready to receive messages.
I then send ten more messages, and see them received in the hw-event-proxy log.
But they do not appear in the consumer.
This test case worked ok with 4.11 and AMQ sidecar.
Reproduce steps:
- deploy SNO cluster with BMER
- Clone https://gitlab.cee.redhat.com/cnf/cnf-gotests tests
- Set BMC_HOSTS to redfish address, sent to developer.
- Set BMC_USER and BMC_PASSWORD
- run the test: FEATURES="ranhwevent" make test-features
- test will configure the application and deploys the consumer
- wait for the consumer to be deployed
- connect to the consumer logs, you can use this:
oc logs -n openshift-bare-metal-events $(oc get pods -n openshift-bare-metal-events|awk '/consumer.*Running/ {print $1}') -f
- wait for the test to restart the hw-event-proxy pod
- when the test sends events again to the new pod
- See that the consumer gets no new events after the new pod is created.
Note:
A workaround is to destroy the consumer and recreate it and then new events arrive.
- clones
-
OCPBUGS-2832 Missing events in consumer after restarting hw-event-proxy with HTTP transport
- Closed
- is blocked by
-
OCPBUGS-2832 Missing events in consumer after restarting hw-event-proxy with HTTP transport
- Closed
- is cloned by
-
OCPBUGS-10210 Default TransportHost to HTTP in 4.12 and 4.13 when transportHost is not specified
- Closed
- links to
- mentioned on