Details
-
Improvement
-
Status: Resolved
-
Major
-
Resolution: Fixed
-
None
-
None
Description
IGNITE-19692 defines a mechanism for executing operations on each partition of a table in a resilient way. It should be implemented so that we can build an index and switch it to an active state when it's built on all partitions. This will allow implement what IGNITE-18539 states:
when index is created, we also create a table somewhere that looks similar to this: indexCompletion.<indexId>.0 = false ... indexCompletion.<indexId>.N = false
This part has not been implemented, and it's separated into another issue.
Things to consider
- RAFT replica listener node may die after sending last "build index" command, thus not updating, so new replica listener must compare states of indexes in schema with states of indexes in storage.
Attachments
Issue Links
- blocks
-
IGNITE-20637 Implement recovery of distributed index building
- Resolved
-
IGNITE-20638 Integration of distributed index building
- Resolved
- causes
-
IGNITE-20636 Add to the MakeIndexAvailableCommand the ability to use only the indexId
- Resolved
- is blocked by
-
IGNITE-20620 Add index availability command to catalog
- Resolved
-
IGNITE-20621 Add a replication group index build completion listener to IndexBuilder
- Resolved
- links to