ui: update changefeed metrics page

[samiskin]

Resolves #98085
Resolves #98088
Resolves #99409
Resolves #100640
Resolves #97931
Resolves #98086

This PR does the following changes:

• Added a scale parameter to Metrics so that I could support a duration metric that's being emitted in Seconds rather than Nanoseconds. Would like frontend feedback on whether this is ok.
• Added support for minutes and hours in Duration graphs
• There is now a "Changefeed Status" graph to show counts of Running/Paused/Failed
• There is now a "Commit Latency" graph to show P50,P90, and P99 commit latencies
• Sink Byte Traffic is now Emitted Bytes
• Sink Timings has been removed because I don't believe either of the metrics exist anymore
• Max Changefeed Latency is now Max Checkpoint Latency
• There is now a Backfill Pending Ranges graph
• There is now a Protected Timestamp Max Age graph
• There is now a Schema Registry Registrations graph

Release note (ui change): The metrics page for changefeeds has been updated with new graphs to track backfill progress, protected timestamps age, and number of schema registry registrations.

[cockroach-teamcity]

This change is 

[shermanCRL]

So great! Should we take the opportunity to add some explanatory text next to each of the graphs?

[miretskiy]

do we want to include number of currently running/paused changefeeds?

[jayshrivastava]

We can close this one too! #97931

[samiskin]

@shermanCRL I have tooltip text for "Max Checkpoint Latency", "Protected Timestamp Max Age", and "Backfill Pending Ranges", and I think the others are pretty self explanatory.

@miretskiy Added at the very top

[miretskiy]

How difficult is it to change the scale on those graphs? Max protected age -- would be nice not to divide by 1e9;
I don't think we need more than a minute granularity.

[samiskin]

Age gets shown as Seconds right now, I looked into adding minutes/hours and it's doable just needs new frontend changes since they currently only support up to seconds. I'll make it show "1h 2m 3s" and see what they think about it.

[shermanCRL]

@samiskin Cool though I think a bit of text saying why you care about this, and how to interpret it -- does it represent a problem -- may help. Maybe a link to relevant docs? I’m imagining that it might prevent an escalation or get a user to a solution faster. @amruss what do you think?

[amruss]

I think that is a good idea, there is a public facing runbook that details why we care about certain changefeed metrics we can either copy from or link to: https://github.com/cockroachlabs/cockroachdb-runbook-template/blob/main/monitoring-alerts/monitoring-dashboard-custom.md#essential-monitoring-metrics-for-changefeeds

[samiskin]

Added P50/P90/P99 histogram for Commit Latency as well as the 2nd graph right below "Changefeed Status"

[samiskin]

Added support for minutes and hours

[dhartunian]

nice! The latest screenshot example looks great. my comments are just around some docs/tests.

Reviewed 6 of 7 files at r1, all commit messages.
Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @HonoreDB and @samiskin)

---

pkg/ui/workspaces/cluster-ui/src/graphs/utils/domain.ts line 224 at r1 (raw file):

}

function ComputeDurationAxisDomain(extent: Extent): AxisDomain {

can you write a few tests for this?

---

pkg/ui/workspaces/db-console/src/views/shared/components/metricQuery/index.tsx line 92 at r1 (raw file):

  title?: string;
  rate?: boolean;
  scale?: number;

I know the existing code doesn't do this but would you mind adding documentation for this parameter? Seems like others might be motivated to use it.

[samiskin]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @dhartunian and @HonoreDB)

---

pkg/ui/workspaces/cluster-ui/src/graphs/utils/domain.ts line 224 at r1 (raw file):

Previously, dhartunian (David Hartunian) wrote…

can you write a few tests for this?

Added tests in a new pkg/ui/workspaces/cluster-ui/src/graphs/utils/domain.spec.ts file 👍

---

pkg/ui/workspaces/db-console/src/views/shared/components/metricQuery/index.tsx line 92 at r1 (raw file):

Previously, dhartunian (David Hartunian) wrote…

I know the existing code doesn't do this but would you mind adding documentation for this parameter? Seems like others might be motivated to use it.

Added comment 👍

[jayshrivastava]

Reviewable status: complete! 0 of 0 LGTMs obtained (waiting on @dhartunian, @HonoreDB, and @samiskin)

a discussion (no related file):
This is really great work :)

---

---

pkg/ccl/changefeedccl/metrics.go line 386 at r3 (raw file):

	metaChangefeedMaxBehindNanos = metric.Metadata{
		Name:        "changefeed.max_behind_nanos",
		Help:        "Time since the changefeed last checkpointed its progress",

The comment implies that it's the time since the most recent checkpoint. It shows the time since the oldest checkpoint. It's also calculated across all changefeeds.

edit: I just saw the tooltip text. We can probably use the first sentence of that here.

---

pkg/ui/workspaces/db-console/src/views/cluster/util/graphs.ts line 62 at r3 (raw file):

      const scaledValues = result.datapoints.map(v => ({
        ...v,
        value: v.value && v.value * (s.props.scale ?? 1),

The && is confusing. Is that right?

[shermanCRL]

pkg/ui/workspaces/db-console/src/views/cluster/util/graphs.ts line 62 at r3 (raw file):

Previously, jayshrivastava (Jayant) wrote…

The && is confusing. Is that right?

It’s kinda idiomatic truthy javascript. Maybe some parens would make it clearer. Or make it more clearly booleans instead of truthy if it’s confusing.

[shermanCRL]

Remind me @samiskin, is Commit Latency the commit-to-emit latency? Asking because, now that we have reduced latency, where is the best place to observe it?

[samiskin]

@shermanCRL Yep, its the time that we confirmed emitting of the message minus the MVCC timestamp. We should be able to view the reduction with this metric, most directly by running without batching (which would add its own latency)

[samiskin]

bors r+

[craig]

🕐 Waiting for PR status (GitHub check) to be set, probably by CI. Bors will automatically try to run when all required PR statuses are set.

[shermanCRL]

Looks like this needs follow-up?

[samiskin]

bors r+

[craig]

Build succeeded:

• Bazel Essential CI (Cockroach)