Replicas shouldn't update top level rem-set on final field delete #745
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Outdated suggestions cannot be applied.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.
This applies to hashmaps, sorted sets and sets.
Consider the following example:
If a sorted set in Dynomite has 2 or more items across replicas,
but if one of the replicas has a missed update(s) and therefore has
only 1 item, removing that item using dynomite previously updated
the top level REM set in that replica.
This is incorrect because now trying to read the other (non-deleted)
field, would cause read repairs to compare its timestamp against the
TS in the top level REM set in the inconsistent replica, and delete
the item incorrectly thinking that someone removed the entire map/set/set
in that broken replica.
This patch fixes it by not updating the top level REM set for field
deletes. Only a DEL will update the top level REM set for maps/sets/zsets.
This was noticed in a TEST cluster.