refactor(executor): construct the executor metrics scope once

[davidlin20dev]

Tracking issue

Closes #7456

Why are the changes needed?

executor/setup.go constructed promutils.NewScope("executor") twice: once for the
webhook setup and once for the plugin setup context. Two independent scopes sharing
the same executor: prefix are a latent duplicate-registration panic on the default
Prometheus registry. If both ever register a like-named metric, prometheus.Register
returns AlreadyRegisteredError and the MustNew* helpers panic at startup.

What changes were proposed in this pull request?

• Construct the base executor scope once (executorScope := promutils.NewScope("executor")).
• Give the webhook and plugin distinct sub-scopes derived from it:
  executorScope.NewSubScope("webhook") and executorScope.NewSubScope("plugin")
  (prefixes executor:webhook: and executor:plugin:).

Scoped to webhook + plugin per the issue. executor:storage and executor:catalog
already use distinct names, so they're collision-safe and left untouched. I'm happy to
fold those into the same base scope (emitted names unchanged) here or in a follow-up
if you'd prefer the file fully consistent.

How was this patch tested?

No new unit test, since this is setup wiring with no new logic, and the issue's
criteria only require the executor to start cleanly and existing tests to pass.
Verified locally with gofmt -l executor/setup.go, go build ./executor/..., and
go vet ./executor/... (all clean). CI exercises ./executor/....

Labels

• changed

Setup process

N/A

Screenshots

N/A

Check all the applicable boxes

• I updated the documentation accordingly.
• All new and existing tests passed.
• All commits are signed-off.

Related PRs

N/A

Docs link

N/A

[pingsutw]

LGTM, thank you!