Description of problem: 'oc adm must-gather -- some_script' should exit with same non-zero code as the failed 'some_script' exits. It should be like 'oc rsh', 'oc rsh POD some_script' exits with same exit code as 'some_script'. Version-Release number of selected component (if applicable): $ oc version Client Version: 4.10.0-202201122238.p0.gda3f635.assembly.stream-da3f635 How reproducible: Always Steps to Reproduce: 1. $ oc adm must-gather -- /usr/bin/gather_audit_logs ... [must-gather-l696h] POD 2022-01-14T10:05:18.217810996Z ERROR: To raise a Red Hat support request, it is required to set the top level audit policy to ... [must-gather-l696h] POD 2022-01-14T10:05:18.217810996Z ... Try 'oc edit apiservers' and set 'spec.audit.profile' back to [must-gather-l696h] POD 2022-01-14T10:05:18.217810996Z "Default" ... ... $ echo $? 0 As per https://github.com/openshift/enhancements/blob/master/enhancements/kube-apiserver/audit-policy.md says: " the command will return with an error (non-zero exit code) " '/usr/bin/gather_audit_logs' indeed exits with 1, but the whole 'oc adm must-gather -- /usr/bin/gather_audit_logs' exits with 0. 2. Try other test: $ oc adm must-gather -- invalid_script $ echo $? 0 Actual results: The whole command 'oc adm must-gather -- some_script' always exits with 0, even if the 'some_script' exits non-zero code. Expected results: The whole command 'oc adm must-gather -- some_script' should exit with same code as 'some_script'. Additional info: Checked/confirmed with Dev in Slack, so filing this bug.
[root@localhost home]# oc adm must-gather -- invalid_script [must-gather ] OUT Using must-gather plug-in image: quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:2fce3fe73a0c94d396f2162d4db64c02fca64238fc0642aca03c6f302df4e916 When opening a support case, bugzilla, or issue please include the following summary data along with any other requested information: ClusterID: ebd965f8-f9db-4b4f-b630-b43fd928748b ClusterVersion: Stable at "4.11.0-0.nightly-2022-05-25-193227" ClusterOperators: clusteroperator/kube-apiserver is not upgradeable because FeatureGatesUpgradeable: "TechPreviewNoUpgrade" does not allow updates [must-gather ] OUT namespace/openshift-must-gather-hdqjl created [must-gather ] OUT clusterrolebinding.rbac.authorization.k8s.io/must-gather-rfp4p created Warning: would violate PodSecurity "restricted:latest": allowPrivilegeEscalation != false (containers "gather", "copy" must set securityContext.allowPrivilegeEscalation=false), unrestricted capabilities (containers "gather", "copy" must set securityContext.capabilities.drop=["ALL"]), runAsNonRoot != true (pod or containers "gather", "copy" must set securityContext.runAsNonRoot=true), seccompProfile (pod or containers "gather", "copy" must set securityContext.seccompProfile.type to "RuntimeDefault" or "Localhost") [must-gather ] OUT pod for plug-in image quay.io/openshift-release-dev/ocp-v4.0-art-dev@sha256:2fce3fe73a0c94d396f2162d4db64c02fca64238fc0642aca03c6f302df4e916 created [must-gather-brfqk] POD 2022-05-31T08:41:47.512178191Z /bin/bash: invalid_script: command not found [must-gather-brfqk] OUT waiting for gather to complete [must-gather-brfqk] OUT downloading gather output [must-gather-brfqk] OUT receiving incremental file list [must-gather-brfqk] OUT ./ [must-gather-brfqk] OUT [must-gather-brfqk] OUT sent 27 bytes received 40 bytes 19.14 bytes/sec [must-gather-brfqk] OUT total size is 0 speedup is 0.00 [must-gather ] OUT namespace/openshift-must-gather-hdqjl deleted [must-gather ] OUT clusterrolebinding.rbac.authorization.k8s.io/must-gather-rfp4p deleted Reprinting Cluster State: When opening a support case, bugzilla, or issue please include the following summary data along with any other requested information: ClusterID: ebd965f8-f9db-4b4f-b630-b43fd928748b ClusterVersion: Stable at "4.11.0-0.nightly-2022-05-25-193227" ClusterOperators: clusteroperator/kube-apiserver is not upgradeable because FeatureGatesUpgradeable: "TechPreviewNoUpgrade" does not allow updates [root@localhost home]# echo $? 0 oc version --client -o yaml clientVersion: buildDate: "2022-05-25T15:55:27Z" compiler: gc gitCommit: 1e9c2f992e8dc18985d789ad4094af685f42d2db gitTreeState: clean gitVersion: 4.11.0-202205251441.p0.g1e9c2f9.assembly.stream-1e9c2f9 goVersion: go1.18 major: "" minor: "" platform: linux/amd64 kustomizeVersion: v4.5.4 releaseClientVersion: 4.11.0-0.nightly-2022-05-26-013101 git log 1e9c2f992e8dc18985d789ad4094af685f42d2db commit 1e9c2f992e8dc18985d789ad4094af685f42d2db Merge: 2e3d4acc9 99913a6cb Author: OpenShift Merge Robot <openshift-merge-robot.github.com> Date: Wed May 25 07:30:46 2022 -0400 Merge pull request #1150 from ardaguclu/bug-2040654 Bug 2040654: Pass pod exit error codes to user
Hi, Attached PR to this bug does not handle cases where script is not found like tested in here; oc adm must-gather -- invalid_script instead it mirrors the exit code when script is valid and returns some exit code other than 0. For example, I tested with this; oc adm must-gather -- exit 123 echo $? 123
Refer https://bugzilla.redhat.com/show_bug.cgi?id=2040654#c3, move this to verified status.
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 (Important: OpenShift Container Platform 4.11.0 bug fix and security update), 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/RHSA-2022:5069