[RFC] Remove quartz/clock from silence and nflog, and drop the dependency

[ultrotter]

This is only used by tests, and we can instead use testing/synctest to get stable mocked time and time.Sleep for instant time advancement in testing/synctest.

Add synctest to TestSilencesSnapshot as it needs stable times even if it doesn't advance them

Pull Request Checklist

Please check all the applicable boxes.

• Is this a Bugfix?
  • Fixes Elimitate "quartz" and use normal golang time #4964
• Is this a breaking change?
  • My changes do not break the existing cluster messages
  • My changes do not break the existing api
• I have signed-off my commits
• I will follow best practices for contributing to this project

Which user-facing changes does this PR introduce?

NONE

[coderabbitai]

📝 Walkthrough

Walkthrough

Removes the quartz clock dependency from silence and nflog code, updates benchmarks and tests to use real wall-clock time, and adds a test.SyncTestRun helper for synchronized subtests.

Changes

Quartz clock removal and synctest migration

Layer / File(s)	Summary
SyncTestRun test helper test/util.go	Adds SyncTestRun, which wraps t.Run in synctest.Test.
Production clock removal go.mod, silence/silence.go, nflog/nflog.go	Removes the quartz dependency and clock fields, and switches time handling to time.Now, time.NewTicker, and time.Since.
Benchmark updates to use real time silence/silence_bench_test.go	Replaces quartz mock clocks with time.Now across benchmarks and removes the internal clock assignment in benchmarkGC.
GC and snapshot tests converted to real time silence/silence_test.go	Updates GC, snapshot, and maintenance tests to use real timestamps, time.Sleep, synctest.Test, and test.SyncTestRun.
Silence set/state tests converted to real time silence/silence_test.go	Refactors silence set, active-silence, failure, and query-state tests to use synctest.Test, s.nowUTC, and time.Sleep.
Expiration, silencer, and annotation tests converted to real time silence/silence_test.go	Converts silence expiration, silencer, log, and annotation tests to synctest.Test with sleep-driven progression.
nflog tests migrated to real time nflog/nflog_test.go	Rewrites nflog tests to remove quartz mocks, use time.Now/time.Sleep, and coordinate maintenance with synctest.

Estimated code review effort: 4 (Complex) | ~45 minutes

Sequence Diagram(s)

sequenceDiagram
  participant Maintenance
  participant Ticker as time.Ticker
  participant Log as nflog.Log
  participant Silences

  Maintenance->>Ticker: time.NewTicker(interval)
  Ticker-->>Maintenance: tick
  Maintenance->>Silences: run maintenance
  Maintenance->>Log: time.Since(start) for completion log

Loading

Related Issues: None specified.

Related PRs: None specified.

Suggested labels: testing, refactor, dependencies

Suggested reviewers: None specified.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Title check	✅ Passed	The title is concise and accurately summarizes removing quartz/clock usage and dropping the dependency in silence and nflog.
Description check	✅ Passed	The description follows the template reasonably well, includes checklist items, an issue reference, and the release-notes block.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

🧹 Nitpick comments (1)

silence/silence_bench_test.go (1)

140-140: 🎯 Functional Correctness | 🔵 Trivial | 💤 Low value

Inconsistent expiry-window margins after switching to real time.

benchmarkMutes/benchmarkMutesParallel had their EndsAt windows bumped from 1 minute to 24 hours when moving off the mock clock (Lines 95, 107, 511), presumably to avoid silences expiring mid-run under real wall-clock time. BenchmarkMutesIncremental, benchmarkQuery, benchmarkQueryParallel, and benchmarkQueryWithConcurrentAdds follow the identical pattern (compute now once, then loop with b.Loop()/b.RunParallel() for the benchtime duration) but were left with only a 1-hour window (Lines 163, 173, 216, 227, 282, 347, 426, 461). With real time now driving expiry, a sufficiently long -benchtime/-count run (e.g., CI stability runs) could let these silences actually expire mid-benchmark, silently skewing QState(SilenceStateActive) results.

Consider extending these windows to match the 24-hour margin applied elsewhere for consistency and to remove this residual flakiness risk.

Also applies to: 163-163, 173-173, 216-216, 227-227, 261-261, 282-282, 328-328, 347-347, 407-407, 426-426, 461-461

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@silence/silence_bench_test.go` at line 140, Extend the silence expiry windows
in the benchmark helpers that use a single cached time value, since
`BenchmarkMutesIncremental`, `benchmarkQuery`, `benchmarkQueryParallel`, and
`benchmarkQueryWithConcurrentAdds` still use a 1-hour `EndsAt` margin while the
real-time benchmarks elsewhere were widened to 24 hours. Update the `EndsAt`
calculations in those benchmark setup blocks to match the 24-hour buffer used by
`benchmarkMutes`/`benchmarkMutesParallel`, keeping the existing `now :=
time.Now()` pattern intact so `QState(SilenceStateActive)` remains stable across
long benchmark runs.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@silence/silence_bench_test.go`:
- Line 140: Extend the silence expiry windows in the benchmark helpers that use
a single cached time value, since `BenchmarkMutesIncremental`, `benchmarkQuery`,
`benchmarkQueryParallel`, and `benchmarkQueryWithConcurrentAdds` still use a
1-hour `EndsAt` margin while the real-time benchmarks elsewhere were widened to
24 hours. Update the `EndsAt` calculations in those benchmark setup blocks to
match the 24-hour buffer used by `benchmarkMutes`/`benchmarkMutesParallel`,
keeping the existing `now := time.Now()` pattern intact so
`QState(SilenceStateActive)` remains stable across long benchmark runs.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 0fd80bec-27c6-4118-ac50-c6645df8d5ae

📥 Commits

Reviewing files that changed from the base of the PR and between 963ed72 and 9b1ad02.

📒 Files selected for processing (4)

• silence/silence.go
• silence/silence_bench_test.go
• silence/silence_test.go
• test/util.go