chore: IBM MQ mixin modernization #1557

schmikei · 2025-11-25T21:19:54Z

Updates the IBM MQ mixin to use more modern libraries

IBM MQ cluster overview

IBM MQ queue manager overview

IBM MQ queue overview

IBM MQ topic overview

I couldn't figure out how to quickly generate topic metrics so I mostly validated query via git diff

IBM MQ logs

Metrics should be flowing to the shared Grafana instance on Nov 25 12pm-5pm EST :)

Dasomeone

Couple minor comments here, overall looks great

ibm-mq-mixin/rows.libsonnet

ibm-mq-mixin/panels.libsonnet

Dasomeone · 2025-12-04T14:53:54Z

ibm-mq-mixin/panels.libsonnet

+      */
+
+      topicMessagesReceived:
+        commonlib.panels.generic.timeSeries.base.new('Topic messages received', targets=[


Bit of a recurring thing for the topics, but I think having an override here for "no data" showing 0 instead might be valid. If you've concerns I'm happy to hear them too, not 100% set on it

Wait I just scrolled a bit further in the screenshots, and I noticed that the subscription status table does in fact show two topics:

dev/telemetry/metrics

dev/telemetry/+

Should these not show up under the topic queries?

I'm assuming that my topic traffic that I tried to simulate did not generate load for these metrics... as noted in the PR description mostly validated via the git diff mostly for these panels since I couldn't quickly get metrics loaded. Let me know if you want me to look deeper into it!

As for the noValue are you suggesting something like this even for a timeSeries:

I suppose I could also do an or vector(0) for those panels? I think it makes sense for stat panels maybe though?

Ah no I specifically meant using the no value standard options override

We should try to avoid using or vector(0) outside of testing as it causes a lot of visualisation artefacts

Also in terms of testing that makes sense, I'd missed the git-dff part for the topic!, all good, I know that's a hard one to generate load on. if you're confident the timeseries panels matches previous (as git diff) then I'm happy with it

ibm-mq-mixin/panels.libsonnet

…modernization

aalhour

A few minor changes.

aalhour · 2025-12-11T14:34:32Z

ibm-mq-mixin/alerts.libsonnet

+                summary: 'There are expired messages, which imply that application resilience is failing.',
+                description:
+                  (
+                    'The number of expired messages in the {{$labels.qmgr}} is {{$labels.value}} which is above the threshold of %(alertsExpiredMessages)s.'


Shouldn't the description reference the $value of metric? Probably need to format it as well: {{ printf "%.0f" $value }}, WDYT?

aalhour · 2025-12-11T14:35:06Z

ibm-mq-mixin/alerts.libsonnet

+                summary: 'Stale messages have been detected.',
+                description:
+                  (
+                    'A stale message with an age of {{$labels.value}} has been sitting in the {{$labels.queue}} which is above the threshold of %(alertsStaleMessagesSeconds)ss.'


Same comment about referencing {{ $value }} instead of {{ $labels.value }}.

aalhour · 2025-12-11T14:35:15Z

ibm-mq-mixin/alerts.libsonnet

+                summary: 'There is limited disk available for a queue manager.',
+                description:
+                  (
+                    'The amount of disk space available for {{$labels.qmgr}} is at {{$labels.value}}%% which is below the threshold of %(alertsLowDiskSpace)s%%.'


Same comment about referencing {{ $value }} instead of {{ $labels.value }}.

aalhour · 2025-12-11T14:35:18Z

ibm-mq-mixin/alerts.libsonnet

+                summary: 'There is a high CPU usage estimate for a queue manager.',
+                description:
+                  (
+                    'The amount of CPU usage for the queue manager {{$labels.qmgr}} is at {{$labels.value}}%% which is above the threshold of %(alertsHighQueueManagerCpuUsage)s%%.'


Same comment about referencing {{ $value }} instead of {{ $labels.value }}.

aalhour · 2025-12-11T14:36:01Z

ibm-mq-mixin/signals/queue.libsonnet

+      name: 'Time on queue',
+      type: 'gauge',
+      description: 'The average time messages spent on the queue.',
+      unit: 'µs',


The metric says _seconds but the unit says micro seconds, can you please change it to 's'?

I do think this one needs updating though, thanks for the catch 👍

aalhour · 2025-12-11T14:52:08Z

ibm-mq-mixin/panels.libsonnet

+          signals.queueManager.queueOperationsMqput.asTarget(),
+        ])
+        + g.panel.timeSeries.panelOptions.withDescription('The number of queue operations of the queue manager.')
+        + g.panel.timeSeries.standardOptions.withUnit('operations')


I don't think 'operations' is a unit that we support in jsonnet-lib/common-lib. Can we use 'short' instead?

aalhour · 2025-12-11T14:52:13Z

ibm-mq-mixin/panels.libsonnet

+          signals.queue.mqputMqput1Count.asTarget(),
+        ])
+        + g.panel.timeSeries.panelOptions.withDescription('The number of queue operations of the queue manager.')
+        + g.panel.timeSeries.standardOptions.withUnit('operations')


I don't think 'operations' is a unit that we support in jsonnet-lib/common-lib. Can we use 'short' instead?

aalhour · 2025-12-11T14:53:28Z

ibm-mq-mixin/signals/queue.libsonnet

Several signals have the unit 'operations' and I don't think it is a supported unit in jsonnet-libs/common-lib. Can we use 'short' instead?

aalhour · 2025-12-11T14:54:02Z

ibm-mq-mixin/signals/queue-manager.libsonnet

Similar to queue signals.

Several signals have the unit 'operations' and I don't think it is a supported unit in jsonnet-libs/common-lib. Can we use 'short' instead?

aalhour · 2025-12-11T14:54:29Z

ibm-mq-mixin/signals/cluster.libsonnet

Similar to the two other signals files.

Several signals have the unit 'operations' and I don't think it is a supported unit in jsonnet-libs/common-lib. Can we use 'short' instead?

operations is preferable here actually. If Grafana doesn't recognise a unit, it's treated as a custom string unit, so it'll be just fine

Dasomeone

Couple more comments, overall I'm happy with it!

ibm-mq-mixin/alerts.libsonnet

Dasomeone · 2025-12-17T15:40:05Z

ibm-mq-mixin/alerts.libsonnet

+            {
+              alert: 'IBMMQLowDiskSpace',
+              expr: |||
+                sum without (description,hostname,instance,job,platform) (ibmmq_qmgr_queue_manager_file_system_free_space_percentage{%(filteringSelector)s}) <= %(alertsLowDiskSpace)s


Suggested change

sum without (description,hostname,instance,job,platform) (ibmmq_qmgr_queue_manager_file_system_free_space_percentage{%(filteringSelector)s}) <= %(alertsLowDiskSpace)s

sum without (description,hostname,instance,job,platform) (ibmmq_qmgr_queue_manager_file_system_free_space_percentage{%(filteringSelector)s}) < %(alertsLowDiskSpace)s

Dasomeone · 2025-12-17T15:44:03Z

ibm-mq-mixin/rows.libsonnet

+
+

Suggested change

Dasomeone · 2025-12-17T15:44:09Z

ibm-mq-mixin/rows.libsonnet

+
+

Suggested change

Dasomeone · 2025-12-17T15:48:43Z

ibm-mq-mixin/panels.libsonnet

+      */
+
+      topicMessagesReceived:
+        commonlib.panels.generic.timeSeries.base.new('Topic messages received', targets=[


Ah no I specifically meant using the no value standard options override

We should try to avoid using or vector(0) outside of testing as it causes a lot of visualisation artefacts

Also in terms of testing that makes sense, I'd missed the git-dff part for the topic!, all good, I know that's a hard one to generate load on. if you're confident the timeseries panels matches previous (as git diff) then I'm happy with it

Dasomeone · 2025-12-17T15:52:41Z

ibm-mq-mixin/signals/cluster.libsonnet

operations is preferable here actually. If Grafana doesn't recognise a unit, it's treated as a custom string unit, so it'll be just fine

Co-authored-by: Emily <[email protected]>

…n topic panels

modernize the IBM MQ mixin

321be7d

schmikei force-pushed the ibm-mq-modernization branch from b01900a to 321be7d Compare November 25, 2025 21:29

schmikei added the monitoring-mixins label Nov 25, 2025

schmikei marked this pull request as ready for review November 25, 2025 21:31

schmikei requested a review from a team as a code owner November 25, 2025 21:31

schmikei mentioned this pull request Nov 25, 2025

chore: Modernize ibm-mq-mixin #1531

Closed

Dasomeone self-assigned this Dec 4, 2025

Dasomeone reviewed Dec 4, 2025

View reviewed changes

schmikei added 5 commits December 9, 2025 10:38

respond to PR feedback; see comments for further clarification questions

83d817d

Merge branch 'master' of github.com:grafana/jsonnet-libs into ibm-mq-…

edad50e

…modernization

empty

86be170

there was a cached dependency locally

cebec52

add name display labels to queue operations pie chart

0ebd4b2

aalhour requested changes Dec 11, 2025

View reviewed changes

aalhour self-assigned this Dec 11, 2025

schmikei added 4 commits December 16, 2025 16:01

pr feedback

0bd869e

use shorts as units

7dc6a17

running into issues with filteringSelector being blank

810779a

merge from main

82c45f7

Dasomeone reviewed Dec 17, 2025

View reviewed changes

schmikei and others added 3 commits December 17, 2025 11:35

Update ibm-mq-mixin/alerts.libsonnet

b2a2a92

Co-authored-by: Emily <[email protected]>

pr feedback; short => operations; alerts thresholds; no values => 0 o…

4a8d296

…n topic panels

Merge branch 'master' into ibm-mq-modernization

b07d93d

	sum without (description,hostname,instance,job,platform) (ibmmq_qmgr_queue_manager_file_system_free_space_percentage{%(filteringSelector)s}) <= %(alertsLowDiskSpace)s
	sum without (description,hostname,instance,job,platform) (ibmmq_qmgr_queue_manager_file_system_free_space_percentage{%(filteringSelector)s}) < %(alertsLowDiskSpace)s

chore: IBM MQ mixin modernization #1557

Are you sure you want to change the base?

chore: IBM MQ mixin modernization #1557

Uh oh!

Conversation

schmikei commented Nov 25, 2025

Uh oh!

Dasomeone left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

aalhour left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Dasomeone left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants