[
https://issues.jboss.org/browse/SWSQE-200?page=com.atlassian.jira.plugin....
]
Kevin Earls commented on SWSQE-200:
-----------------------------------
Hi [~gbaufake] Sorry, I'm not sure I understand your comments? Have you updated the
b12 cluster or should I be running somewhere else now?
If we're staying on B12, the first step to determining whether the cluster is working
correctly would be to do the following:
-- Run this job from the Jenkins in the jaeger-test project multiple times:
https://jenkins-jaeger-test.openshift3.jonqe.lab.eng.bos.redhat.com/job/F...
-- It should succeed on every run
-- If it fails, it will leave a jaeger all-in-one instance running. The instance will
look ok in the OpenShift console, but if you click on the jaeger-query link you will get
"Application is not available"
At the moment the job fails on every second iteration. If we can get past this error then
we can work towards resolving the ElasticSearch problem.
B12 OpenShift Cluster is flaky
------------------------------
Key: SWSQE-200
URL:
https://issues.jboss.org/browse/SWSQE-200
Project: Kiali QE
Issue Type: QE Task
Reporter: Kevin Earls
Assignee: Guilherme Baufaker RĂªgo
Priority: Blocker
I'm opening this mostly as a placeholder, and will update it as I get more
information on the problems I've been experiencing. Since I've been using B12
I've been experiencing more frequent failures than on other clusters, including
minishift on my laptop and the CNCF CI Jenkins. Here are a couple of instances:
1. Deploying ElasticSearch along with the Jaeger Production templates requires
allocating 2G or memory, even though the default 512M works fine elsewhere. This can be
seen in the Jaeger Standalone Performance tests job here:
https://jenkins-jaeger-test.openshift3.jonqe.lab.eng.bos.redhat.com/job/J...
. If you run the job without changing the ES_MEMORY parameter to 2Gi it will fail.
2. I have a set of smoke tests for Red Hat productized artifacts for the Jaeger Java
client which are run using the Jaeger all-in-one template. It's fairly simple, but on
B12 deployment of the Jaeger all-in-one images fails on every other build. I've made
a copy of the job here:
https://jenkins-jaeger-test.openshift3.jonqe.lab.eng.bos.redhat.com/job/F... and
its history should be clear.
So far I have not been able to get any useful information about why this is failing. In
the OpenShift console it looks like Jaeger has started correctly. There are no errors in
the logs, nor any to be found under monitoring. But if you click on the Jaeger link, you
get the message "Application is not available . The application is currently not
serving requests at this endpoint. It may not have been started or is still
starting."
--
This message was sent by Atlassian JIRA
(v7.5.0#75005)