Skip to content

HDDS-15656. Fix non-atomic data size accumulation in ContainerBalancer move callback #10708

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
sreejasahithi wants to merge 1 commit into
base: master
Choose a base branch
from
+13 −29
Open

HDDS-15656. Fix non-atomic data size accumulation in ContainerBalancer move callback #10708

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
42 changes: 13 additions & 29 deletions ...cm/src/main/java/org/apache/hadoop/hdds/scm/container/balancer/ContainerBalancerTask.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -89,8 +89,6 @@ public class ContainerBalancerTask implements Runnable {
private long maxSizeToMovePerIteration;
private int countDatanodesInvolvedPerIteration;
private long sizeScheduledForMoveInLatestIteration;
// count actual size moved in bytes
private long sizeActuallyMovedInLatestIteration;
private final List<DatanodeUsageInfo> overUtilizedNodes;
private final List<DatanodeUsageInfo> underUtilizedNodes;
private Set<String> excludeNodes;
Expand Down Expand Up @@ -338,29 +336,18 @@ private ContainerBalancerTaskIterationStatusInfo getIterationStatistic(Integer i
ContainerMoveInfo containerMoveInfo = new ContainerMoveInfo(metrics);

DataMoveInfo dataMoveInfo =
getDataMoveInfo(currentIterationResultName, sizeEnteringDataToNodes, sizeLeavingDataFromNodes);
getDataMoveInfo(sizeEnteringDataToNodes, sizeLeavingDataFromNodes);
return new ContainerBalancerTaskIterationStatusInfo(iterationInfo, containerMoveInfo, dataMoveInfo);
}

private DataMoveInfo getDataMoveInfo(String currentIterationResultName, Map<DatanodeID, Long> sizeEnteringDataToNodes,
private DataMoveInfo getDataMoveInfo(Map<DatanodeID, Long> sizeEnteringDataToNodes,
Map<DatanodeID, Long> sizeLeavingDataFromNodes) {
if (currentIterationResultName == null) {
// For unfinished iteration
return new DataMoveInfo(
getSizeScheduledForMoveInLatestIteration(),
sizeActuallyMovedInLatestIteration,
sizeEnteringDataToNodes,
sizeLeavingDataFromNodes
);
} else {
// For finished iteration
return new DataMoveInfo(
getSizeScheduledForMoveInLatestIteration(),
metrics.getDataSizeMovedInLatestIteration(),
sizeEnteringDataToNodes,
sizeLeavingDataFromNodes
);
}
return new DataMoveInfo(
getSizeScheduledForMoveInLatestIteration(),
metrics.getDataSizeMovedInLatestIteration(),
sizeEnteringDataToNodes,
sizeLeavingDataFromNodes
);
}

private Map<DatanodeID, Long> convertToNodeIdToTrafficMap(Map<DatanodeDetails, Long> nodeTrafficMap) {
Expand Down Expand Up @@ -743,9 +730,8 @@ private void checkIterationMoveResults() {

metrics.incrementNumContainerMovesTimeout(metrics.getNumContainerMovesTimeoutInLatestIteration());

metrics.incrementDataSizeMovedGBInLatestIteration(sizeActuallyMovedInLatestIteration / OzoneConsts.GB);

metrics.incrementDataSizeMovedInLatestIteration(sizeActuallyMovedInLatestIteration);
long bytesMovedInLatestIteration = metrics.getDataSizeMovedInLatestIteration();
metrics.incrementDataSizeMovedGBInLatestIteration(bytesMovedInLatestIteration / OzoneConsts.GB);

metrics.incrementDataSizeMovedGB(metrics.getDataSizeMovedGBInLatestIteration());

Expand All @@ -754,8 +740,8 @@ private void checkIterationMoveResults() {
LOG.info("Iteration Summary. Number of Datanodes involved: {}. Size " +
"moved: {} ({} Bytes). Number of Container moves completed: {}.",
countDatanodesInvolvedPerIteration,
byteDesc(sizeActuallyMovedInLatestIteration),
sizeActuallyMovedInLatestIteration,
byteDesc(bytesMovedInLatestIteration),
bytesMovedInLatestIteration,
metrics.getNumContainerMovesCompletedInLatestIteration());
}

Expand Down Expand Up @@ -943,8 +929,7 @@ private boolean moveContainer(DatanodeDetails source,
metrics.incrementNumContainerMovesFailedInLatestIteration(1);
} else {
if (result == MoveManager.MoveResult.COMPLETED) {
sizeActuallyMovedInLatestIteration +=
containerInfo.getUsedBytes();
metrics.incrementDataSizeMovedInLatestIteration(containerInfo.getUsedBytes());
LOG.debug("Container move completed for container {} from " +
"source {} to target {}", containerID, source,
moveSelection.getTargetNode());
Expand Down Expand Up @@ -1156,7 +1141,6 @@ private void resetState() {
this.selectedTargets.clear();
this.countDatanodesInvolvedPerIteration = 0;
this.sizeScheduledForMoveInLatestIteration = 0;
this.sizeActuallyMovedInLatestIteration = 0;
metrics.resetDataSizeMovedGBInLatestIteration();
metrics.resetDataSizeMovedInLatestIteration();
metrics.resetNumContainerMovesScheduledInLatestIteration();
Expand Down