sanpwc commented on code in PR #2475:
URL: https://github.com/apache/ignite-3/pull/2475#discussion_r1301971938
##########
modules/table/src/main/java/org/apache/ignite/internal/table/distributed/raft/PartitionListener.java:
##########
@@ -94,7 +94,10 @@ public class PartitionListener implements RaftGroupListener {
/** Storage of transaction metadata. */
private final TxStateStorage txStateStorage;
- /** Rows that were inserted, updated or removed. All row IDs are sorted in
natural order to prevent deadlocks upon commit/abort. */
+ /**
+ * Rows that were inserted, updated or removed. All row IDs are sorted in
natural order to prevent deadlocks upon commit/abort.
+ * TODO: IGNITE-18502 Implement a pending update storage
Review Comment:
What do you mean? Why do we need to implement pending update storage and how
it's related with removing full partition scan on node restart?
The basic idea that we will substitute explicit scan with writeIntent
resolution with further clean-up where writeIntent has all required data to
perfrom the resolution, meaning that txsPendingRowIds is not needed for that
process. It is however needed for common TX clean-up.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
