Report exemplars for histogram and timer metrics

[sfackler]

Before this PR

We didn't report any exemplars for metrics, making it a bit harder to investigate slowness or other badness.

After this PR

==COMMIT_MSG==
Histogram and timer metrics now report exemplars.
==COMMIT_MSG==

The behavior here matches WC-Java - we report at most a single exemplar per metric, corresponding to the measurement from the sampled trace made within the last reporting window which had the highest value (i.e. was the slowest for a timer metric).

Depends on palantir/witchcraft-rust-logging#40.

Metric output with an exemplar:

{"type":"metric.1","time":"2025-11-30T22:06:08.389198Z","metricName":"server.response","metricType":"timer","values":{"1m":0.024078182461403894,"count":2,"max":270.667,"p95":270.667,"p99":270.667,"p999":270.667},"samples":[{"value":270.667,"time":"2025-11-30T22:05:49.878164Z","traceId":"aa9ab66b2e728b7a"}],"tags":{"endpoint":"foo","service-name":"TestResource"}}

[changelog-app]

Generate changelog in changelog/@unreleased

What do the change types mean?

• feature: A new feature of the service.
• improvement: An incremental improvement in the functionality or operation of the service.
• fix: Remedies the incorrect behaviour of a component of the service in a backwards-compatible way.
• break: Has the potential to break consumers of this service's API, inclusive of both Palantir services
  and external consumers of the service's API (e.g. customer-written software or integrations).
• deprecation: Advertises the intention to remove service functionality without any change to the
  operation of the service itself.
• manualTask: Requires the possibility of manual intervention (running a script, eyeballing configuration,
  performing database surgery, ...) at the time of upgrade for it to succeed.
• migration: A fully automatic upgrade migration task with no engineer input required.

Note: only one type should be chosen.

How are new versions calculated?

• ❗The break and manual task changelog types will result in a major release!
• 🐛 The fix changelog type will result in a minor release in most cases, and a patch release version for patch branches. This behaviour is configurable in autorelease.
• ✨ All others will result in a minor version release.

Type

• Feature
• Improvement
• Fix
• Break
• Deprecation
• Manual task
• Migration

Description

Histogram and timer metrics now report exemplars.

Check the box to generate changelog(s)

• Generate changelog entry

[stale]

This PR has been automatically marked as stale because it has not been touched in the last 14 days. If you'd like to keep it open, please leave a comment or add the 'long-lived' label, otherwise it'll be closed in 7 days.

[stale]

This PR has been automatically marked as stale because it has not been touched in the last 14 days. If you'd like to keep it open, please leave a comment or add the 'long-lived' label, otherwise it'll be closed in 7 days.

[sfackler]

These are required to ensure that the zipkin thread local state is set when the endpoint metric layer handles its timer updates.

[sfackler]

Drive-by fix - previously the max would be rounded down to the nearest whole microsecond while the percentiles wouldn't.