fix buffered key below watermark

[morgo]

A Pull Request should be associated with an Issue.

We wish to have discussions in Issues. A single issue may be targeted by multiple PRs.
If you're offering a new feature or fixing anything, we'd like to know beforehand in Issues,
and potentially we'll be able to point development in a particular direction.
Further notes in https://github.com/block/spirit/blob/main/.github/CONTRIBUTING.md

Fixes #624

[Copilot]

Pull request overview

This PR fixes a critical logic bug in the bufferedMap subscription where the watermark optimization check was inverted. The bug caused keys below the low watermark (already copied by the copier) to be incorrectly skipped during flush operations, while keys at or above the watermark (not yet copied) were incorrectly flushed. This could lead to data race conditions, contention, and performance/memory issues from buffering too many versions.

Changes:

• Fixed the watermark check condition in bufferedMap.Flush() by adding a negation operator to match deltaMap behavior
• Updated test expectations in TestBufferedMapFlushWithoutLockRespectsWatermark to verify the correct behavior
• Updated comments in the test to accurately describe the fixed watermark logic

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
pkg/repl/subscription_buffered.go	Fixed inverted watermark logic by changing condition from KeyBelowLowWatermark(...) to !KeyBelowLowWatermark(...) and updated debug log message
pkg/repl/subscription_buffered_test.go	Updated test expectations and comments to reflect the corrected watermark behavior, verifying that keys below the watermark are now flushed and keys at/above are skipped

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Kiran01bm]

lgtm - also good that normal migrations (without buffering) was not impacted by this bug.