[unsafePtr]

HeartbeatConsistencyCheckPingsAsync asserted minExpected <= opCount && opCount >= minExpected, checking the minimum twice and never the maxExpected from the theory data. It now checks opCount <= maxExpected, so the case with consistency checks disabled verifies that no heartbeat operations are sent.

Both theory cases pass on net8.0 and net10.0 against a local Redis 7, three runs in a row.

Checklist

• I fully and freely contribute this code in accordance with the project license (and am legally able to do so)
• I take responsibility for this contribution's quality and correctness, including any portions produced with AI assistance (see CONTRIBUTING.md).

[mgravell]

Thanks @unsafePtr — nice catch. The typo dates all the way back to #2656, so the upper bound (and the false case's exact-zero) had been silently vacuous for three years; the compound assert testing minExpected twice is exactly the shape that hides this, which is a good argument for Assert.InRange here.

One thing we'll keep an eye on post-merge: since those bounds have never actually gated a run, this effectively arms two new timing assertions (the 100ms heartbeat fires on wall clock, so a Task.Delay(1s) overshoot under CI load can push past 15 pings, and the false case now requires literally zero ops in the window). If they flake we'll derive the ceiling from measured elapsed time rather than loosening the constants — but that's on us, not a blocker for this fix. Merging — thanks again!

[unsafePtr]

@mgravell , that's the beauty of roslyn analyzer. Let me try to scan couple more projects. If you have also something in mind for catching such kind of mistakes, please let me know. I would be happy to improve the analyzer.
https://github.com/unsafePtr/Typo.Analyzers