[Debian-ha-maintainers] Bug#1040165: pacemaker: Migration regression after upgrading bullseye to bookworm

[Florent Carli]

Hello,

I could not reproduce with the packages you mentionned, so it's a good sign.
The version shown in crm status is version 2.1.5-a3f44794f94.

root at virtu-elabo2:~# dpkg -l | grep -E
"(libcib|libcrm|liblrmd|libpe-|libstonit|pacemaker)"
ii  crmsh                                   4.4.1-1
            all          CRM shell for the pacemaker cluster manager
ii  libcib27:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager CIB library
ii  libcrmcluster29:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager cluster library
ii  libcrmcommon34:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager common library
ii  libcrmservice28:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager service library
ii  liblrmd28:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager LRMD library
ii  libpacemaker1:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager utility library
ii  libpe-rules26:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager Policy Engine rules library
ii  libpe-status28:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager Policy Engine status library
ii  libstonith1                             1.0.12-21+b2
            amd64        Reusable cluster libraries -- libstonith1
ii  libstonithd26:amd64
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager STONITH daemon library
ii  pacemaker
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager
ii  pacemaker-cli-utils
2.1.5-1+salsaci+20230702+262           amd64        cluster resource
manager command line utilities
ii  pacemaker-common
2.1.5-1+salsaci+20230702+262           all          cluster resource
manager common files
ii  pacemaker-resource-agents
2.1.5-1+salsaci+20230702+262           all          cluster resource
manager general resource agents

root at virtu-elabo2:~# crm status
Status of pacemakerd: 'Pacemaker is running' (last updated 2023-07-03
14:55:59 +02:00)
Cluster Summary:
  * Stack: corosync
  * Current DC: virtu-elabo2 (version 2.1.5-a3f44794f94) - partition with quorum
  * Last updated: Mon Jul  3 14:56:00 2023
  * Last change:  Mon Jul  3 09:50:25 2023 by root via crm_resource on
virtu-elabo2
  * 3 nodes configured
  * 13 resource instances configured (1 DISABLED)

Node List:
  * Online: [ virtu-elabo1 virtu-elabo2 virtu-elabo3 ]

Full List of Resources:
  * Clone Set: cl_ping_test [ping_test]:
    * Started: [ virtu-elabo1 virtu-elabo2 virtu-elabo3 ]
  * st-mms    (stonith:external/iec61580_mms):     Started virtu-elabo2
  * st-ssh    (stonith:external/ssh-virtu):     Started virtu-elabo1
  * ABB    (ocf:seapath:VirtualDomain):     Stopped (disabled)
  * Clone Set: cl_ptpstatus_test [ptpstatus_test]:
    * Started: [ virtu-elabo1 virtu-elabo2 virtu-elabo3 ]
  * debian    (ocf:seapath:VirtualDomain):     Started virtu-elabo2
  * debian2    (ocf:seapath:VirtualDomain):     Started virtu-elabo2
  * debian3    (ocf:seapath:VirtualDomain):     Started virtu-elabo2
  * TESTVMU    (ocf:seapath:VirtualDomain):     Started virtu-elabo1

Thanks.
Florent.

On Sun, Jul 2, 2023 at 10:08 PM Ferenc Wágner <wferi at niif.hu> wrote:
>
> Hi,
>
> I've pushed the proposed backported fix to
> https://salsa.debian.org/ha-team/pacemaker/-/commits/wferi/debian/bookworm/,
> it would be great if you could test it.  You can download the build
> artifacts from https://salsa.debian.org/ha-team/pacemaker/-/pipelines/547159
> if you don't want to build the packages yourselves.  If they don't
> exhibit the bug, I'll ask for a stable update.
> --
> Regards,
> Feri.