ENG-2835 websocket monitoring daemon (endpoint-monitor)

[zhwrd]

Description

Adds websocket monitoring daemon for edge-proxyd and downstream infra providers. This first iteration is pretty specific to one method: eth_subscribe. I opted to re-use the eth client instead of rolling my own websocket client since I understand we mainly only allow the subscribe method... Also in general, think we'd be better off using a vendor for most synthetic testing on our edge proxies since it will let us ping the endpoint from several regions without having to deploy there...

In a loop, runs eth_subscribe over websocket for each configured provider and reports metrics to prometheus on any errors encountered.

Metadata

• ENG-2835

[changeset-bot]

⚠️ No Changeset found

Latest commit: c77b42e

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[mergify]

This PR changes implementation code, but doesn't include a changeset. Did you forget to add one?

[mslipper]

Looks good. A couple nits:

• We generally use urfave/cli elsewhere in the codebase to do env var parsing/CLI management. It's nice because it lets you specify command line args as well. It also integrates with the op-service package, which contains generalized utilities for passing in metric configs and serving them.
• We also use op-service to create instances of the Prometheus metrics collector rather than serving a global. See https://github.com/ethereum-optimism/optimism/tree/develop/op-service/metrics for the package in question.
• We usually do manual signal handling so we can gracefully shut things down on SIGINT.

These aren't blocking, though, given the scope + goals of this project.

Lastly, this PR is under the size SLA when you discount the go.sum file.

[zhwrd]

Looks good. A couple nits:

• We generally use urfave/cli elsewhere in the codebase to do env var parsing/CLI management. It's nice because it lets you specify command line args as well. It also integrates with the op-service package, which contains generalized utilities for passing in metric configs and serving them.
• We also use op-service to create instances of the Prometheus metrics collector rather than serving a global. See https://github.com/ethereum-optimism/optimism/tree/develop/op-service/metrics for the package in question.
• We usually do manual signal handling so we can gracefully shut things down on SIGINT.

These aren't blocking, though, given the scope + goals of this project.

Lastly, this PR is under the size SLA when you discount the go.sum file.

cool, will go ahead and adopt the op-service standard stuff, looks useful!

[zhwrd]

Think this looks good now... Don't think i can gracefully shutdown the metrics server in this case because don't have access to the underlying Server from the op-server/metrics package but correct me if i'm wrong? Anyways it's not really important here so all good.

[mslipper]

Do these envconfig tags carry meaning anymore?