[IMPROVED] Optimize catchups for replicas and mirrors with large interior deletes. #4929
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Under scenarios where interior deletes were orders of magnitude larger than the msgs themselves, the system would need to transfer all interior deletes as skip msgs. This change adopts the ability to send DeleteRanges on streamSyncRequests, and for R1 mirrors to locally use the new store.SkipMsgs to optimize speed.
This CL also has a bug fix for memory store state with delete details, and several general improvements to the filestore impl.
Signed-off-by: Derek Collison derek@nats.io