openinx commented on a change in pull request #437: HBASE-22539 WAL corruption
due to early DBBs re-use when Durability.A…
URL: https://github.com/apache/hbase/pull/437#discussion_r309958049
##########
File path:
hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/wal/AsyncFSWAL.java
##########
@@ -323,7 +324,9 @@ private void syncFailed(long epochWhenSync, Throwable
error) {
private void syncCompleted(AsyncWriter writer, long processedTxid, long
startTimeNs) {
highestSyncedTxid.set(processedTxid);
for (Iterator<FSWALEntry> iter = unackedAppends.iterator();
iter.hasNext();) {
- if (iter.next().getTxid() <= processedTxid) {
+ FSWALEntry entry = iter.next();
+ if (entry.getTxid() <= processedTxid) {
+ entry.getRpcCall().ifPresent(ServerCall::releaseByWAL);
Review comment:
Still worry about the CPU-consuming in ServerCall#release, for AsyncFSWAL,
we will accumulate the batch size entries and then call the sync, so if the
sync not finished, then the ServerCall#release will be always looping..
----------------------------------------------------------------
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]
With regards,
Apache Git Services
