Kill AOF BulkConsume When Connection is Dead

[vazois]

This PR introduces a stopConsume callback that checks the health of the client connection when streaming the AOF.
I used a callback here to avoid maintaining a second background task that monitors the connection and kills the AOF BulkConsume loop.

[Copilot]

Pull request overview

This PR aims to stop the AOF streaming BulkConsumeAllAsync loop when the replica client connection is no longer healthy, by introducing a stopConsume callback that can terminate consumption without running a separate monitoring task.

Changes:

• Added an optional stopConsume callback parameter to TsavoriteLogScanIterator.BulkConsumeAllAsync.
• Wired stopConsume into the primary-side AOF sync task to stop streaming when GarnetClientSession disconnects.
• Updated the replica replay task call site to accommodate the new method signature.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 5 comments.

File	Description
libs/storage/Tsavorite/cs/src/core/TsavoriteLog/TsavoriteLogIterator.cs	Adds stopConsume callback support to the bulk consume loop.
libs/cluster/Server/Replication/PrimaryOps/AofSyncTaskInfo.cs	Uses stopConsume to exit AOF streaming when the client disconnects.
libs/cluster/Server/Replication/ReplicaOps/ReplicaReplayTask.cs	Updates BulkConsumeAllAsync invocation to match the new signature.

---

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