Fix arangorestore API to create _rev on shard leader [BTS-1232]. #18191
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Scope & Purpose
This is about https://arangodb.atlassian.net/browse/BTS-1232 .
Short description:
arangorestore uses the (undocumented) API /_api/replication/restore-data
implemented in the RestReplicationHandler. Since ArangoDB 3.8, we
create new _rev values during the restore operation. Unfortunately,
this creation of _rev values was done on the coordinator(s) and since
these are hybrid logical clock (HLC) values, it can happen that the
generated _rev values on two different coordinators are the same for
two different documents. This will clash on the leader of the OneShard
database, since there, the newly generated _rev values are used as
internal document IDs, which directly translate into a RocksDB key in
the DocumentsColumnFamily. This leads to a collision and this manifests
itself as a lock timeout, since two different RocksDB transactions try
to touch the same key. However, even if no lock timeout would have
happened, the consequences would have been dire: Namely, there would
have been a corruption between primary index and data in the Documents
column family. We would have had two primary index entries pointing to
the same document.
Checklist
Related Information