[ClusterLabs] [EXT] Re: Re: Q: Configuring ldirectord in SLES15

Windl, Ulrich u.windl at ukr.de
Wed Jun 25 13:01:47 UTC 2025


Hi!

Some steps further:
I was wondering why ldirectord would terminate on all cluster nodes until I realized that my checkcommand failed to check the status actually.
However, as it turned out, my check program would write an error message to stderr, but ldirectord swallows that, so it never ended up somewhere. The problem was that $0 in “external-perl” had a value that was not expected when writing the check program; so eventually I had to hard-code the path for my configuration file into the check program.
Maybe see also https://serverfault.com/q/1187240/407952

Kind regards,
Ulrich Windl

From: Xin Liang <xliang at suse.com>
Sent: Tuesday, June 24, 2025 3:21 PM
To: Windl, Ulrich <u.windl at ukr.de>
Cc: Cluster Labs - All topics related to open-source clustering welcomed <users at clusterlabs.org>
Subject: [EXT] Re: Re: [ClusterLabs] Q: Configuring ldirectord in SLES15

Hi,

Please see `crm ra help info`,
```
info [<class>:[<provider>:]]<type>
```
This is the required format of `crm ra info` command, not related to how it is displayed in crm_mon

On Tue, Jun 24, 2025 at 9:01 PM Windl, Ulrich <u.windl at ukr.de<mailto:u.windl at ukr.de>> wrote:

From: Users <users-bounces at clusterlabs.org<mailto:users-bounces at clusterlabs.org>> On Behalf Of Xin Liang via Users
Sent: Tuesday, June 24, 2025 10:32 AM
To: Cluster Labs - All topics related to open-source clustering welcomed <users at clusterlabs.org<mailto:users at clusterlabs.org>>
Cc: Xin Liang <xliang at suse.com<mailto:xliang at suse.com>>
Subject: [EXT] Re: [ClusterLabs] Q: Configuring ldirectord in SLES15

Hi Ulrich,

Please try `crm ra info ocf:heartbeat:ldirectord`
[Windl, Ulrich]
Oops, as crm_mon had displayed “(ocf::heartbeat:ldirectord)” in the old cluster I had tried it with double-colon, but that failed.
So the heartbeat variant is still supported (with the preferred config file located in /etc/ha.d/)?

Is the double-colon a long-standing bug in crm_mon, or is it a bug in crm shell?

Kind regards,
Ulrich

On Tue, Jun 24, 2025 at 3:27 PM Windl, Ulrich <u.windl at ukr.de<mailto:u.windl at ukr.de>> wrote:
Hi!

While trying to migrate our ldirector configuration from SLES12 to SLES15 I noticed that the “RA” (which is a systemd service now) does not have any parameters, while in SLES12 I still could use “params configfile=…”. So in SLES12 I had a separate ldirectord for each service, using a separate configuration file.
How is this intended to work in SLES15 (SP5 or newer)? I see that SLES seemed to have dropped most Ras in favor of using systemd services, but IMHO systemd services are quite a different thing.

I see that in SLES12 the RA is in /usr/lib/ocf/resource.d/heartbeat/ldirectord (resource-agents-4.3.018.a7fb5035-3.107.2), but the CRM shell does not offer that when completing for “ra info”.
The file is still there in SLES15 SP5 (resource-agents-4.12.0+git30.7fd7c8fa-150500.3.15.3), but still it’s not offered by “crm ra info”. Why? A bug?

If I try “bute force”, then I get:
# crm ra info heartbeat:ldirectord
ERROR: heartbeat:ldirectord: got no meta-data, does this RA exist?

Kind regards,
Ulrich Windl

_______________________________________________
Manage your subscription:
https://lists.clusterlabs.org/mailman/listinfo/users

ClusterLabs home: https://www.clusterlabs.org/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.clusterlabs.org/pipermail/users/attachments/20250625/fc05eeec/attachment.htm>


More information about the Users mailing list