Bug 1408476
| Summary: | pcs does not currently work with systemd's instantiated services | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Michele Baldessari <michele> | ||||||
| Component: | pcs | Assignee: | Ivan Devat <idevat> | ||||||
| Status: | CLOSED ERRATA | QA Contact: | cluster-qe <cluster-qe> | ||||||
| Severity: | urgent | Docs Contact: | |||||||
| Priority: | urgent | ||||||||
| Version: | 7.3 | CC: | cfeist, cluster-maint, gfidente, idevat, mlisik, omular, rsteiger, tojeline | ||||||
| Target Milestone: | rc | Keywords: | ZStream | ||||||
| Target Release: | --- | ||||||||
| Hardware: | All | ||||||||
| OS: | Linux | ||||||||
| Whiteboard: | |||||||||
| Fixed In Version: | pcs-0.9.156-1.el7 | Doc Type: | Bug Fix | ||||||
| Doc Text: |
Cause:
User creates a cluster resource based on a systemd service which supports multiple instances (i.e. there is the "@" character in the resource agent name).
Consequence:
The resource agent is not recognized by pcs and the command fails. User has to specify --force in the command for it to succeed.
Fix:
Fixed validation of systemd resource agent names.
Result:
It is possible to create cluster resource based on a systemd service which supports multiple instances.
|
Story Points: | --- | ||||||
| Clone Of: | |||||||||
| : | 1412244 (view as bug list) | Environment: | |||||||
| Last Closed: | 2017-08-01 18:26:07 UTC | Type: | Bug | ||||||
| Regression: | --- | Mount Type: | --- | ||||||
| Documentation: | --- | CRM: | |||||||
| Verified Versions: | Category: | --- | |||||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||||
| Embargoed: | |||||||||
| Bug Depends On: | |||||||||
| Bug Blocks: | 1335682, 1412244 | ||||||||
| Attachments: |
|
||||||||
Created attachment 1239460 [details]
proposed fix
This will be eventually fixed by a rebase. Until then (and for the z-stream) here is the patch.
Giulio, could you provide us some feedback on this just to be 100% sure we covered your use case correctly? hi guys, thanks a lot! I've tested manually the "proposed fix" attachment together with https://review.openstack.org/#/c/407039/ (meant to deploy the ceph rbd-mirror daemon) and it completed successfully: [root@overcloud-controller-0 ~]# pcs status|grep -i rbd ceph-rbd-mirror_openstack (systemd:ceph-rbd-mirror@openstack): Started overcloud-controller-0 looking forward to use it in OSPd11 :) After Fix: [vm-rhel72-1 ~] $ rpm -q pcs pcs-0.9.156-1.el7.x86_64 [vm-rhel72-1 ~] $ pcs resource create gettytest systemd:getty@tty3 [vm-rhel72-1 ~] $ pcs status|grep gettytest gettytest (systemd:getty@tty3): Started vm-rhel72-2 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, 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-2017:1958 |
Created attachment 1235077 [details] Locally tested fix Description of problem: Currently (pcs-0.9.152-10.el7.centos.x86_64) we do not support systemd instantiated services (0pointer.de/blog/projects/instances.html) aka the services that have an @ in the service filename and that can be instantiated multiple times. For example: [root@centos pcs]# pcs resource create gettytest systemd:getty@tty3 Error: Unable to create resource 'systemd:getty@tty3', it is not installed on this system (use --force to override) With --force this does get created and it works, but pcs is not considering the instantiated service scenario when it does the prevalidation. I attach the trivial patch that fixes it for me. I attach it here to the downstream bug only, because this is actually already fixed in pcs master via the following change: commit 491ac087fd06480a9fda0dfc5dd97f6b9dfbae56 Author: Tomas Jelinek <tojeline> Date: Tue Oct 11 17:12:11 2016 +0200 get resource/fence agents info from pacemaker With the attached patch applied I can correctly do the following: [root@centos system]# pcs resource create gettytest4 systemd:getty@tty4 [root@centos system]# pcs status Cluster name: clustername Stack: corosync Current DC: centos (version 1.1.15-11.el7_3.2-e174ec8) - partition with quorum Last updated: Fri Dec 23 16:40:53 2016 Last change: Fri Dec 23 16:40:49 2016 by root via cibadmin on centos 1 node and 4 resources configured Online: [ centos ] Full list of resources: gettytest4 (systemd:getty@tty4): Started centos