Document missing metrics [CTT-942]

[JamesHazelcast]

Various metrics were identified as missing in our documentation - this PR adds those metrics for released functionality.

Fixes https://hazelcast.atlassian.net/browse/CTT-942

[netlify]

✅ Deploy Preview for hardcore-allen-f5257d ready!

Name	Link
🔨 Latest commit	9847ead
🔍 Latest deploy log	https://app.netlify.com/projects/hardcore-allen-f5257d/deploys/6998903c77da0d0008655b5f
😎 Deploy Preview	https://deploy-preview-2091--hardcore-allen-f5257d.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[Rob-Hazelcast]

LGTM, thanks!

[anatoly-hazelcast]

I wonder whether this PR is intended to cover all the missing metrics. There are around 50 metrics in the initial list but less than 20 are covered here

[JamesHazelcast]

I wonder whether this PR is intended to cover all the missing metrics. There are around 50 metrics in the initial list but less than 20 are covered here

The original list looks like it handles things like query cache somewhat oddly - perhaps an artifact of renaming for Prometheus? Those sort of metrics are already covered in docs, so I didn't duplicate them.

Are there any particular metrics missing that you can't find elsewhere @anatoly-hazelcast?

[anatoly-hazelcast]

@JamesHazelcast I've dooublechecked and indeed some of the metrics from the original list seem to be coming from the latest Platform and ARE already documented, e.g. map_index_averageHitLatency. So, at least a subset of those shouldn't appear in this PR

However e.g. the metrics prefixed with hz_map_nearcache_ are coming from the com.hazelcast.internal.metrics.MetricDescriptorConstants#MAP_PREFIX_NEARCACHE and have nothing to do with any kind of renaming.
The stuff like hz_memorymanager_ is coming from com.hazelcast.internal.metrics.MetricDescriptorConstants#CLIENT_PREFIX_MEMORY_MANAGER
the hz_operation_operationTimeoutCount comes from com.hazelcast.internal.metrics.MetricDescriptorConstants#OPERATION_METRIC_OPERATION_SERVICE_TIMEOUT_COUNT
And I neither can't find these metrics in the existing docs https://docs.hazelcast.com/hazelcast/5.7-snapshot/list-of-metrics nor in this PR
There's more examples but I'm struggling to understand how these examples I've listed above could be the artifacts of the Prometheus-renaming if they are coming straight from the Platform.

[JamesHazelcast]

@anatoly-hazelcast here's a list of all the metrics from your original Slack message, and their status: https://hazelcast.atlassian.net/browse/CTT-942?focusedCommentId=141420

As you can see, all of the metric values are now covered in documentation - the only open question is about some of the prefix differences; i.e. map.nearcache vs nearcache, and probably more important the memory vs memorymanager vs memorymanager.stats (which are all separate memory systems, but use the same value metrics). @Rob-Hazelcast how would you want these duplicate metric values handled in docs? If we had all unique keys listed, we'd be duplicating the metrics descriptors for memory at least 3x (and making 3x the maintenance ofc).

[Rob-Hazelcast]

@anatoly-hazelcast here's a list of all the metrics from your original Slack message, and their status: https://hazelcast.atlassian.net/browse/CTT-942?focusedCommentId=141420

As you can see, all of the metric values are now covered in documentation - the only open question is about some of the prefix differences; i.e. map.nearcache vs nearcache, and probably more important the memory vs memorymanager vs memorymanager.stats (which are all separate memory systems, but use the same value metrics). @Rob-Hazelcast how would you want these duplicate metric values handled in docs? If we had all unique keys listed, we'd be duplicating the metrics descriptors for memory at least 3x (and making 3x the maintenance ofc).

Oof. Can we have an extra column for the variant prefixes?

[JamesHazelcast]

Oof. Can we have an extra column for the variant prefixes?

@Rob-Hazelcast That works, but I think I may have spotted an even bigger gap (or I'm blind) - the list_of_metrics docs page is all about member side metrics. Where do we document client-side metrics? I can't find it - but I don't think we should mix client metrics into this page (so nearcache. would be listed under client metrics, map.nearcache would be under members - but again this might duplicate maintenance requirements).

edit: I think the page is intended to list all metrics (both client/member), but it's just not clear - I'll prepare some edits.

[anatoly-hazelcast]

I still keep wondering why we're having these variants and more importantly why we're keeping them if they are effectively the same.. But at least now I got the idea of what prefixes we're talking about 😅
Thanks for the clarification in the ticket!

[Rob-Hazelcast]

Thanks James. Clarifications around POV and prefixes look good.

[github-actions]

❌ Failed to backport, change must be manually backported.

[github-actions]

❌ Failed to backport, change must be manually backported.

[github-actions]

👍 Created #2110 to backport into v/5.6.