Hide Forgot
Description of problem: The output of `crictl ps --help` includes a `--pod` or `-p` option to filter a container listing by Pod ID: ---- [core@master-0 ~]$ sudo crictl ps --help NAME: crictl ps - List containers USAGE: crictl ps [command options] [arguments...] OPTIONS: <etc> --pod value, -p value Filter by pod id <etc> ---- However, the option returns no results when run on OpenShift Container Platform (OCP) 4.5 CoreOS nodes. Version-Release number of selected component (if applicable): OCP 4.5 CoreOS nodes running . How reproducible: Every time. Steps to Reproduce: 1. Check cluster and CoreOS versions: ---- [core@master-0 ~]$ oc get clusterversion NAME VERSION AVAILABLE PROGRESSING SINCE STATUS version 4.5.2 True False 35m Cluster version is 4.5.2 [core@master-0 ~]$ oc debug node/<string>-master-0 Starting pod/<string->master-0-debug ... To use host binaries, run `chroot /host` Pod IP: <IP_address> If you don't see a command prompt, try pressing enter. sh-4.2# chroot /host sh-4.4# cat /etc/redhat-release Red Hat Enterprise Linux CoreOS release 4.5 ---- 2. Obtain a Pod ID: ---- sh-4.4# sudo crictl pods | head -2 POD ID CREATED STATE NAME NAMESPACE ATTEMPT 77d0fcd4ba290 30 seconds ago Ready <string>-master-0-debug default 0 ---- 3. Filter a list of containers by the Pod ID output in the preceding command. None of the following command formats return any containers: ---- sh-4.4# sudo crictl ps --pod 77d0fcd4ba290 CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID ---- ---- sh-4.4# sudo crictl ps --pod=77d0fcd4ba290 CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID ---- ---- sh-4.4# sudo crictl ps -p 77d0fcd4ba290 CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID ---- 4. Running the `crictl ps` command without the `--pod` option and grepping for the same Pod ID returns a container: ---- sh-4.4# sudo crictl ps | grep '77d0fcd4ba290' 5c598a5a0481d <registry> About a minute ago Running container-00 0 77d0fcd4ba290 ---- Actual results: The `-p` or `--pod` option in `crictl ps` does not return any containers. Expected results: For the option to list any containers related to the given Pod ID.
it's not a super awesome user interface, but you need to specify the full pod id. What you see in `crictl pods` is truncated. you can find the full pod id with `crictl inspectp $trunc_id | jq -r .status.id` if filtering based on the full pod id doesn't work, I will look into it. otherwise, I'm inclined to believe this is not a bug (but maybe a crictl feature request?)
(In reply to Peter Hunt from comment #1) > it's not a super awesome user interface, but you need to specify the full > pod id. What you see in `crictl pods` is truncated. you can find the full > pod id with > `crictl inspectp $trunc_id | jq -r .status.id` > > if filtering based on the full pod id doesn't work, I will look into it. > otherwise, I'm inclined to believe this is not a bug (but maybe a crictl > feature request?) Thank you Peter for explaining that. I tested this using the non-truncated Pod ID and it worked on an OCP 4.5 node. My view is: 1. The output of `crictl ps --help` should clarify that the non-truncated Pod ID is required. But, this is still not idea because it is not obvious as to where to find the non-truncated ID without an explanation. 2. For consistency, the `--pod` option should accept the Pod ID as it is output by other `crictl` commands, i.e. truncated. I'm not sure if that is a feature request. In my view the current option is not consistent and it is not obvious in the help menu as to how to use it. Paul.
I have opened a docs pr upstream to solve this
actually, we are changing behavior in CRI-O (attached PR)
upstream PR has merged, fix will be in 4.6
not fixed in version : 4.6.0-0.nightly-2020-08-04-210224 sh-4.4# crictl pods | head -2 POD ID CREATED STATE NAME NAMESPACE ATTEMPT dd778ce2e6163 3 minutes ago Ready ip-10-0-207-124us-east-2computeinternal-debug default 0 sh-4.4# crictl ps -p dd778ce2e6163 CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID sh-4.4# crictl ps -p dd778ce2e6163a804d011985cb95275078f13cd3465d38c391e83ebd4fde6300 (long pod id is ok) CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID 9deeaaab2865d quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:6011b9ece8fa6e259949c98f0a30a05d9bd314e02e364e42723cebeba55c53c6 4 minutes ago Running container-00 0 dd778ce2e6163
ah, this is because the machine-os-content hasn't pulled in the correct version yet. here's the latest machine-os-content: https://releases-rhcos-art.cloud.privileged.psi.redhat.com/contents.html?stream=releases%2Frhcos-4.6&release=46.82.202008032211-0 which has cri-o: 1.19.0-61.rhaos4.6.git79c1228.el8 (https://github.com/cri-o/cri-o/commit/79c122895fe645f6b642cd20ea6eea78cf41d085) cri-o is fixed in https://github.com/cri-o/cri-o/commit/15e88b8f4492d4d6c3348c2f184239f04635c3d3 which is included in https://brewweb.engineering.redhat.com/brew/buildinfo?buildID=1276422
verified in version : 4.6.0-0.nightly-2020-08-06-131904 sh-4.4# crictl version Version: 0.1.0 RuntimeName: cri-o RuntimeVersion: 1.19.0-67.rhaos4.6.gitc964c31.el8-dev RuntimeApiVersion: v1alpha1 sh-4.4# crictl pods | head -2 POD ID CREATED STATE NAME NAMESPACE ATTEMPT 08a6a6ca2572a 2 minutes ago Ready redhat-marketplace-qh8jq openshift-marketplace 0 sh-4.4# crictl ps -p 08a6a6ca2572a CONTAINER IMAGE CREATED STATE NAME ATTEMPT POD ID a86b4cff4c1ea registry.redhat.io/redhat/certified-operator-index@sha256:39b4589658b766142baef3bd20d21114d8d0571e50c5c38d1a13f248c41c1e68 2 minutes ago Running registry-server 0 08a6a6ca2572a
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory (OpenShift Container Platform 4.6 GA Images), and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:4196