Note: This bug is displayed in read-only format because
the product is no longer active in Red Hat Bugzilla.
RHEL Engineering is moving the tracking of its product development work on RHEL 6 through RHEL 9 to Red Hat Jira (issues.redhat.com). If you're a Red Hat customer, please continue to file support cases via the Red Hat customer portal. If you're not, please head to the "RHEL project" in Red Hat Jira and file new tickets here. Individual Bugzilla bugs in the statuses "NEW", "ASSIGNED", and "POST" are being migrated throughout September 2023. Bugs of Red Hat partners with an assigned Engineering Partner Manager (EPM) are migrated in late September as per pre-agreed dates. Bugs against components "kernel", "kernel-rt", and "kpatch" are only migrated if still in "NEW" or "ASSIGNED". If you cannot log in to RH Jira, please consult article #7032570. That failing, please send an e-mail to the RH Jira admins at rh-issues@redhat.com to troubleshoot your issue as a user management inquiry. The email creates a ServiceNow ticket with Red Hat. Individual Bugzilla bugs that are migrated will be moved to status "CLOSED", resolution "MIGRATED", and set with "MigratedToJIRA" in "Keywords". The link to the successor Jira issue will be found under "Links", have a little "two-footprint" icon next to it, and direct you to the "RHEL project" in Red Hat Jira (issue links are of type "https://issues.redhat.com/browse/RHEL-XXXX", where "X" is a digit). This same link will be available in a blue banner at the top of the page informing you that that bug has been migrated.
As of RHEL 8.4 (via Bug 1371576), Pacemaker will support a boolean "influence" option to colocation constraints, indicating whether the dependent resource's location preferences should affect the main resource's placement.
The default of true is the previous behavior. False essentially means that a failed dependent resource should stop rather than cause the main resource to move.
Proposed interface: add influence=true/false to the possible [options] for pcs constraint colocation add/set.
As of the same feature, resources can also take a boolean "critical" meta-attribute to serve as a default for "influence" for all constraints where the resource is the dependent resource. I believe this will not require any changes in pcs.
I just realized that pcs appears to pass constraint options along without restricting them to known names, so nothing special needs to be done to support "influence".
However Marketa Smazova in QE raised the question of whether "critical" should change how a colocation constraint is displayed. In other words, if someone configures a colocation with influence=false, then "pcs constraint colocation show" will display "(influence:false)", but if someone sets critical=false on the resource instead, "pcs constraint colocation show" will not give any indication that influence has been disabled.
I don't have a preference on how that should be handled, but it's a good question to think about.
As Ken mentioned in comment 1, setting 'influence' constraint option and 'critical' resource meta-attribute is already possible with pcs. Displaying '(influence:false)' in colocation constraints works as well.
Displaying '(critical:false)' in collocation constraints next to resources is a different story, though. The meta-attribute can be either set in a resource, in its parent (clones, groups) or in resource defaults. Resource defaults support rules, and resource meta-attributes can also use them, even though pcs is not capable of configuring them yet. In these cases, displaying simply '(critical:false)' would be misleading, as it would not be true all the time, and displaying all the details with rules would be confusing. Displaying an applied configuration is currently not possible either: 1) pcs would have to reimplement pacemaker's logic to determine the applied configuration, 2) 'pcs constraint colocation config' displays configuration and not currently applied values (that would have to be another command), so we don't want to mix those two.
Therefore, I'm closing this as wontfix. If this issue becomes serious, feel free to reopen the bz.