tkalkirill commented on a change in pull request #9001:
URL: https://github.com/apache/ignite/pull/9001#discussion_r613816957
##########
File path:
modules/core/src/main/java/org/apache/ignite/internal/processors/cache/persistence/wal/FileWriteAheadLogManager.java
##########
@@ -1325,6 +1325,14 @@ private FileWriteHandle rollOver(FileWriteHandle cur,
@Nullable WALRecord rec) t
private FileWriteHandle restoreWriteHandle(@Nullable WALPointer
lastReadPtr) throws StorageException {
long absIdx = lastReadPtr == null ? 0 : lastReadPtr.index();
+ FileDescriptor[] walArchiveFiles = walArchiveFiles();
+
+ if (archiver != null &&
+ !F.isEmpty(walArchiveFiles) &&
+ absIdx <= walArchiveFiles[walArchiveFiles.length - 1].idx) {
+ absIdx = walArchiveFiles[walArchiveFiles.length - 1].idx + 1;
Review comment:
@nizhikov
We define a segment into which we will continue to write based on the last
read WAL record. In the normal case, everything works well because after
rollover, we write a record to a new segment, see
**FileWriteAheadLogManager#log**.
In the root cause of the current ticket, we do not write any WAL record to
the new segment, and at the same time archive the previous segment, so after
restarting the node, we will try to read the WAL records in a new segment, but
they are not there, so we will ignore it, so we continue to write to the
previous segment. If we had written a record to the auto-rolloved segment, then
there would be no problem since we would have read this WAL record in a new
segment and continued writing to a new segment.
--
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.
For queries about this service, please contact Infrastructure at:
[email protected]
