fix: disable histogram polling after repeated failures

[100NikhilBro]

Summary

Fix infinite error logging in latency histogram polling for unsupported Redis versions.

Changes

• Added runtimeCapabilityTracker.recordFailure to histogram error handling
• Aligns behavior with latency events handler
• Disables polling after repeated failures

Result

Prevents continuous error logs for Redis/Valkey instances that do not support LATENCY HISTOGRAM.

Closes #43

[100NikhilBro]

Hi @KIvanow , just a gentle follow-up on this PR.

All checks are passing, and I’ve kept the change aligned with the existing latency events handling pattern. Please let me know if any adjustments are needed.

Thanks for your time!

[100NikhilBro]

Hi @KIvanow , just checking in once more on this PR whenever you get a chance.

No rush at all — I just wanted to see if there’s any feedback or changes needed from my side.

Thanks!

[KIvanow]

Hi @100NikhilBro, great catch! And thank you for the contribution!

Could you please sign your commit(s) and then I'll merge the PR.

[100NikhilBro]

Thanks! I’ve signed the commit and force-pushed the updated branch.

It looks like the current failing API test may be unrelated to this PR, since the changes here are limited to latency-analytics.service.ts. Happy to make any adjustments if needed.

[KIvanow]

Thanks! I’ve signed the commit and force-pushed the updated branch.

It looks like the current failing API test may be unrelated to this PR, since the changes here are limited to latency-analytics.service.ts. Happy to make any adjustments if needed.

@100NikhilBro no worries on the API, that was a regression that I caused and fixed in a subsequent commit yesterday, but you synced in between. It's fixed on master/main