Bug 1210833
| Summary: | pcs is not fully distributed (contains local-only set/query operations) | ||
|---|---|---|---|
| Product: | Red Hat Enterprise Linux 7 | Reporter: | Jan Pokorný [poki] <jpokorny> |
| Component: | pcs | Assignee: | Tomas Jelinek <tojeline> |
| Status: | CLOSED WONTFIX | QA Contact: | cluster-qe <cluster-qe> |
| Severity: | unspecified | Docs Contact: | |
| Priority: | unspecified | ||
| Version: | 7.2 | CC: | cfeist, cluster-maint, phagara, tojeline |
| Target Milestone: | rc | ||
| Target Release: | --- | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | If docs needed, set a value | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2020-12-15 07:34:01 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: | |||
This bug was originally triggered by discovery that pcs does not depend
on pacemaker{,-cli} in RHEL 6:
# rpm -qR pcs | grep -e pacemaker -e pc | echo Nothing
> Nothing
# rpm -q pcs
> pcs-0.9.139-5.el6.x86_64
Hence I was expecting that this is another form of saying "install pcs
whenever you want, manage arbitrary cluster with that incl. just the
external ones (i.e., the machine running pcs doesn't need to have
anything in common with clusters modulo pcs itself
[no pacemaker CLI tools need to be installed, etc.])".
re [comment 1]: Apparently: s/whenever/wherever *** Bug 1217166 has been marked as a duplicate of this bug. *** A very specific use-case (commands run on virt-002):
# crm_attribute -t nodes -N virt-002 -n maintenance -v on
OR, when available
# pcs node maintenance virt-002
+
# systemctl status pacemaker.service
# pcs cluster status
> Error: cluster is not currently running on this node
while it (with or without interactive prompt or an extra option set) could
connect to one the active peers (if present) and indeed show the status
for the cluster that virt-002 is still a member of (just temporarily put
into the maintenance mode).
re [comment 5]: s/systemctl status pacemaker.service/systemctl stop pacemaker.service/ After evaluating this issue, there are no plans to address it further or fix it in an upcoming release. Therefore, it is being closed. If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened. |
IMHO it would be appropriate to "distributify", at least by the mean of specifying remote node (and provided that authentication is set up) as opposed to local-only, all set/query operations provided. IIRC, with ccs, you could always specify node(s) you want to work with -- such modus operandi is simply missing in pcs in some cases. "pcs resources list" is an example of the operation violating such expectation. This is strange as pcs can be run on a dedicated machine outside of cluster and hence resource-agents package is likely not installed then. [For this operation specifically, it would even make sense to be sensitive about number of nodes to query - single one: standard output as usual - more: redacted output stating which resources are common, and then per-node extras] There may be more local-only operations to look at.