[ClusterLabs] Why did the ticket that was granted become revoked?

[yuusuke_iida at newson.co.jp]

Hi all,

While testing a multi-site cluster using tickets, I encountered an issue where tickets were revoked automatically when Pacemaker was restarted.

Things to confirm
* Is it intentional that tickets are revoked when Pacemaker is restarted?

What we are trying to do
* We are verifying a multi-site cluster using tickets.
* Since we assume that tickets will be switched manually, we are not using booth.

Expected behavior
* We want tickets to remain granted even after Pacemaker restarts.

What we did
* For maintenance purposes, we stopped Pacemaker on one node with `pcs cluster stop` and then started it with `pcs cluster start`, but the tickets that were in the granted state were revoked.

Best regards,

Cluster status before Pacemaker stop on standby1
```
Cluster name: primary_cluster
Cluster Summary:
  * Stack: corosync (Pacemaker is running)
  * Current DC: primary1 (1) (version 2.1.8-3.el9-3980678f0) - partition with quorum
  * Last updated: Thu Jul 31 23:06:30 2025 on standby1
  * Last change:  Thu Jul 31 23:06:05 2025 by root via root on primary1
  * 2 nodes configured
  * 13 resource instances configured

Node List:
  * Node primary1 (1): online, feature set 3.19.5
  * Node standby1 (2): online, feature set 3.19.5

Full List of Resources:
  * Clone Set: pgsql-clone [pgsql] (promotable):
    * pgsql (ocf:linuxhajp:pgsql):  Promoted primary1
    * pgsql (ocf:linuxhajp:pgsql):  Unpromoted standby1
  * Resource Group: primary-group:
    * ipaddr-primary (ocf:heartbeat:IPaddr2):  Started primary1
    * ipaddr-replication (ocf:heartbeat:IPaddr2):  Started primary1
  * ipaddr-standby (ocf:heartbeat:IPaddr2):  Started standby1
  * Resource Group: remote-site-rep-group:
    * remote-site-pgsql (ocf:linuxhajp:pgsql):  Stopped
  * local-site-pgsql (ocf:linuxhajp:pgsql):  Stopped
  * Clone Set: ping-clone [ping]:
    * ping (ocf:pacemaker:ping):  Started primary1
    * ping (ocf:pacemaker:ping):  Started standby1
  * Clone Set: storage-mon-clone [storage-mon]:
    * storage-mon (ocf:heartbeat:storage-mon):  Started primary1
    * storage-mon (ocf:heartbeat:storage-mon):  Started standby1
  * fence1-virsh (stonith:fence_virsh):  Started standby1
  * fence2-virsh (stonith:fence_virsh):  Started primary1

Node Attributes:
  * Node: primary1 (1):
    * master-pgsql                     : 1000
    * pgsql-data-status                : LATEST
    * pgsql-master-baseline            : 000000002D000168
    * pgsql-status                     : PRI
    * ping-status                      : 1
  * Node: standby1 (2):
    * master-pgsql                     : 100
    * pgsql-data-status                : STREAMING|SYNC
    * pgsql-status                     : HS:sync
    * ping-status                      : 1

Migration Summary:

Fencing History:
  * turning off of standby1 successful: delegate=a human, client=stonith_admin.3019770, origin=primary1, completed='2025-07-31 23:02:10.031118 -04:00'

Tickets:
  tokyo-dc-dr-site-ticket revoked
  tokyo-dc-primary-site-ticket granted  (granted=true, last-granted=Thu Jul 31 23:03:47 2025)

PCSD Status:
  primary1: Online
  standby1: Online

Daemon Status:
  corosync: active/disabled
  pacemaker: active/disabled
  pcsd: active/enabled
```

Cluster status after Pacemaker start on standby1
```
Cluster name: primary_cluster
Cluster Summary:
  * Stack: corosync (Pacemaker is running)
  * Current DC: primary1 (1) (version 2.1.8-3.el9-3980678f0) - partition with quorum
  * Last updated: Thu Jul 31 23:07:08 2025 on standby1
  * Last change:  Thu Jul 31 23:06:36 2025 by root via root on primary1
  * 2 nodes configured
  * 13 resource instances configured

Node List:
  * Node primary1 (1): online, feature set 3.19.5
  * Node standby1 (2): online, feature set 3.19.5

Full List of Resources:
  * Clone Set: pgsql-clone [pgsql] (promotable):
    * pgsql (ocf:linuxhajp:pgsql):  Stopped
    * pgsql (ocf:linuxhajp:pgsql):  Stopped
  * Resource Group: primary-group:
    * ipaddr-primary (ocf:heartbeat:IPaddr2):  Stopped
    * ipaddr-replication (ocf:heartbeat:IPaddr2):  Stopped
  * ipaddr-standby (ocf:heartbeat:IPaddr2):  Stopped
  * Resource Group: remote-site-rep-group:
    * remote-site-pgsql (ocf:linuxhajp:pgsql):  Stopped
  * local-site-pgsql (ocf:linuxhajp:pgsql):  Stopped
  * Clone Set: ping-clone [ping]:
    * ping (ocf:pacemaker:ping):  Started standby1
    * ping (ocf:pacemaker:ping):  Started primary1
  * Clone Set: storage-mon-clone [storage-mon]:
    * storage-mon (ocf:heartbeat:storage-mon):  Started standby1
    * storage-mon (ocf:heartbeat:storage-mon):  Started primary1
  * fence1-virsh (stonith:fence_virsh):  Started standby1
  * fence2-virsh (stonith:fence_virsh):  Started primary1

Node Attributes:
  * Node: primary1 (1):
    * master-pgsql                     : -INFINITY
    * pgsql-data-status                : LATEST
    * pgsql-status                     : STOP
    * ping-status                      : 1
  * Node: standby1 (2):
    * pgsql-data-status                : DISCONNECT
    * ping-status                      : 1

Migration Summary:

Fencing History:
  * turning off of standby1 successful: delegate=a human, client=stonith_admin.3019770, origin=primary1, completed='2025-07-31 23:02:10.031118 -04:00'

Tickets:
  tokyo-dc-dr-site-ticket revoked
  tokyo-dc-primary-site-ticket revoked

PCSD Status:
  primary1: Online
  standby1: Online

Daemon Status:
  corosync: active/disabled
  pacemaker: active/disabled
  pcsd: active/enabled
```
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/users/attachments/20250801/cf680b28/attachment-0001.htm>