Add tokio metrics

[sfackler]

Before this PR

We didn't have any metrics tracking the state of tasks in the Tokio runtime.

After this PR

==COMMIT_MSG==
Added metrics tracking the state of the Tokio runtime.
==COMMIT_MSG==

Unfortunately, many of these rely on unstable Tokio APIs. As a result, you have to opt-in both with the standard tokio_unstable cfg and a tokio_unstable Cargo feature in this crate.

[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

Added metrics tracking the state of the Tokio runtime.

Check the box to generate changelog(s)

• Generate changelog entry

[sfackler]

I am a bit unsure about what values are appropriate here. This setup gives us 20 buckets with these ranges:

0ns..65.536µs
65.536µs..131.072µs
131.072µs..262.144µs
262.144µs..524.288µs
524.288µs..1.048576ms
1.048576ms..2.097152ms
2.097152ms..4.194304ms
4.194304ms..8.388608ms
8.388608ms..16.777216ms
16.777216ms..33.554432ms
33.554432ms..67.108864ms
67.108864ms..134.217728ms
134.217728ms..268.435456ms
268.435456ms..536.870912ms
536.870912ms..1.073741824s
1.073741824s..2.147483648s
2.147483648s..4.294967296s
4.294967296s..8.589934592s
8.589934592s..17.179869184s
17.179869184s..18446744073.709551615s

The intent is to avoid spamming the metrics infrastructure with a huge number of buckets, while still giving us enough information to go off of. Future polls under 100us or so are in a totally good place and I don't think we really care about splitting those out, and above a few seconds the poll is so unreasonably long the specific length doesn't matter too much.

[sfackler]

This is mostly useful to normalize the values of tokio.tasks.poll-duration-bucket into percentages.

[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.