HBASE-29039 Optimize read performance for accumulated delete markers on the same row or cell

[EungsopYoo]

No description provided.

[Apache-HBase]

🎊 +1 overall

Vote	Subsystem	Runtime	Logfile	Comment
+0 🆗	reexec	0m 20s		Docker mode activated.
			_ Prechecks _
+1 💚	dupname	0m 0s		No case conflicting files found.
+0 🆗	codespell	0m 0s		codespell was not available.
+0 🆗	detsecrets	0m 0s		detect-secrets was not available.
+1 💚	@author	0m 0s		The patch does not contain any @author tags.
+1 💚	hbaseanti	0m 0s		Patch does not have any anti-patterns.
			_ master Compile Tests _
+1 💚	mvninstall	3m 51s		master passed
+1 💚	compile	3m 33s		master passed
+1 💚	checkstyle	0m 39s		master passed
+1 💚	spotbugs	1m 50s		master passed
+1 💚	spotless	0m 47s		branch has no errors when running spotless:check.
			_ Patch Compile Tests _
+1 💚	mvninstall	3m 39s		the patch passed
+1 💚	compile	3m 30s		the patch passed
+1 💚	javac	3m 30s		the patch passed
+1 💚	blanks	0m 0s		The patch has no blanks issues.
-0 ⚠️	checkstyle	0m 38s	/results-checkstyle-hbase-server.txt	hbase-server: The patch generated 5 new + 8 unchanged - 0 fixed = 13 total (was 8)
+1 💚	spotbugs	1m 45s		the patch passed
+1 💚	hadoopcheck	15m 22s		Patch does not cause any errors with Hadoop 3.3.6 3.4.0.
+1 💚	spotless	1m 8s		patch has no errors when running spotless:check.
			_ Other Tests _
+1 💚	asflicense	0m 17s		The patch does not generate ASF License warnings.
		46m 37s

Subsystem	Report/Notes
Docker	ClientAPI=1.47 ServerAPI=1.47 base: https://ci-hbase.apache.org/job/HBase-PreCommit-GitHub-PR/job/PR-6557/6/artifact/yetus-general-check/output/Dockerfile
GITHUB PR	#6557
Optional Tests	dupname asflicense javac spotbugs checkstyle codespell detsecrets compile hadoopcheck hbaseanti spotless
uname	Linux 9b584a4c6b5b 5.4.0-200-generic #220-Ubuntu SMP Fri Sep 27 13:19:16 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux
Build tool	maven
Personality	dev-support/hbase-personality.sh
git revision	master / 80044c2
Default Java	Eclipse Adoptium-17.0.11+9
Max. process+thread count	83 (vs. ulimit of 30000)
modules	C: hbase-server U: hbase-server
Console output	https://ci-hbase.apache.org/job/HBase-PreCommit-GitHub-PR/job/PR-6557/6/console
versions	git=2.34.1 maven=3.9.8 spotbugs=4.7.3
Powered by	Apache Yetus 0.15.0 https://yetus.apache.org

This message was automatically generated.

[Apache-HBase]

🎊 +1 overall

Vote	Subsystem	Runtime	Logfile	Comment
+0 🆗	reexec	0m 20s		Docker mode activated.
-0 ⚠️	yetus	0m 2s		Unprocessed flag(s): --brief-report-file --spotbugs-strict-precheck --author-ignore-list --blanks-eol-ignore-file --blanks-tabs-ignore-file --quick-hadoopcheck
			_ Prechecks _
			_ master Compile Tests _
+1 💚	mvninstall	4m 2s		master passed
+1 💚	compile	1m 8s		master passed
+1 💚	javadoc	0m 34s		master passed
+1 💚	shadedjars	6m 7s		branch has no errors when building our shaded downstream artifacts.
			_ Patch Compile Tests _
+1 💚	mvninstall	3m 37s		the patch passed
+1 💚	compile	1m 7s		the patch passed
+1 💚	javac	1m 7s		the patch passed
+1 💚	javadoc	0m 30s		the patch passed
+1 💚	shadedjars	5m 52s		patch has no errors when building our shaded downstream artifacts.
			_ Other Tests _
+1 💚	unit	181m 4s		hbase-server in the patch passed.
		208m 40s

Subsystem	Report/Notes
Docker	ClientAPI=1.47 ServerAPI=1.47 base: https://ci-hbase.apache.org/job/HBase-PreCommit-GitHub-PR/job/PR-6557/6/artifact/yetus-jdk17-hadoop3-check/output/Dockerfile
GITHUB PR	#6557
Optional Tests	javac javadoc unit compile shadedjars
uname	Linux 8aa279d51e08 5.4.0-200-generic #220-Ubuntu SMP Fri Sep 27 13:19:16 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux
Build tool	maven
Personality	dev-support/hbase-personality.sh
git revision	master / 80044c2
Default Java	Eclipse Adoptium-17.0.11+9
Test Results	https://ci-hbase.apache.org/job/HBase-PreCommit-GitHub-PR/job/PR-6557/6/testReport/
Max. process+thread count	5216 (vs. ulimit of 30000)
modules	C: hbase-server U: hbase-server
Console output	https://ci-hbase.apache.org/job/HBase-PreCommit-GitHub-PR/job/PR-6557/6/console
versions	git=2.34.1 maven=3.9.8
Powered by	Apache Yetus 0.15.0 https://yetus.apache.org

This message was automatically generated.

[Apache9]

I checked the code, we do have logic to seek to next row or column when we hit a delte family cell.

hbase/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/querymatcher/ScanQueryMatcher.java

Line 207 in 28c4353

switch (deleteResult) {

But the problem is that, seems we will return earlier before we actually call this method here

hbase/hbase-server/src/main/java/org/apache/hadoop/hbase/regionserver/querymatcher/NormalUserScanQueryMatcher.java

Line 76 in 28c4353

returnCode = checkDeleted(deletes, cell);

The above code block

    if (PrivateCellUtil.isDelete(typeByte)) {
      boolean includeDeleteMarker =
        seePastDeleteMarkers ? tr.withinTimeRange(timestamp) : tr.withinOrAfterTimeRange(timestamp);
      if (includeDeleteMarker) {
        this.deletes.add(cell);
      }
      return MatchCode.SKIP;
    }

Seems incorrect, we will always return MatchCode.SKIP if we get a delete maker...

I think why we do not find this before is that, usually there will be only one delete maker, so when we check the next cell, we will fall through and call the checkDeleted method so we will seek to next row or column.

Here the scenario is that we have bunch of delete makrer, then here we will see them all instead of seek to next row or column, since we will always go into the code block above and return MatchCode.SKIP.

I think we should try to optimize the logic of the above code block.

Thanks.

[virajjasani]

FYI @kadirozde