Chia-Ping Tsai created YUNIKORN-2218:
----------------------------------------
Summary: It always requests apps from nonexistent queues after the
e2e gets failed
Key: YUNIKORN-2218
URL: https://issues.apache.org/jira/browse/YUNIKORN-2218
Project: Apache YuniKorn
Issue Type: Test
Reporter: Chia-Ping Tsai
[https://github.com/apache/yunikorn-k8shim/blob/master/test/e2e/wrappers.go#L112]
{code:go}
for _, ns := range namespaces {
logErr := k8s.LogNamespaceInfo(ns)
Ω(logErr).NotTo(HaveOccurred())
pods, err := k.GetPodsByOptions(metav1.ListOptions{})
Ω(err).NotTo(HaveOccurred())
By("Pod count is " + strconv.Itoa(len(pods.Items)))
for _, pod := range pods.Items {
By("Pod name is " + pod.Name)
By("Pod details: " + pod.String())
}
logErr = restClient.LogAppsInfo(ns)
Ω(logErr).NotTo(HaveOccurred())
}
{code}
`restClient.LogAppsInfo(ns)` tries to get apps from "root.$ns". However, these
queues are nonexistent normally
(https://github.com/apache/yunikorn-k8shim/blob/master/test/e2e/framework/helpers/yunikorn/rest_api_utils.go#L351)
It logged all apps before https://github.com/apache/yunikorn-k8shim/pull/516
got merged. It seems to me previous behavior makes sense. Logging all apps is
better than logging nothing :)
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
