[apache/helix] -- Issue during onboarding resources without instances

himanshukandwal commented 4 months ago

Issues

[x] My PR addresses the following Helix issues and references them in the PR description: Fixes #2781

Description

[x] Here are some details about my PR, including screenshots of any UI changes: When adding a new WAGED resource with a tag and without any instances against that tag, we are observing NPE coming from the system. This in turn fails the complete WAGED cluster rebalance pipeline. This particularly happen when the numPartitions = 0.

To solve this issue we are adding a check in the ResourceComputationStage to have such resources excluded from the pipeline computation and only be considered when there are actual resource partitions (>0) to be assigned to the instances.

org.apache.helix.HelixRebalanceException: Failed to compute for delayed rebalance overwrites in cluster ZnRecord=CLUSTER_TestWagedClusterExpansionWithAddingResourcesBeforeInstances, {DELAY_REBALANCE_ENABLED=true, DELAY_REBALANCE_TIME=3000000, FAULT_ZONE_TYPE=zone, PERSIST_BEST_POSSIBLE_ASSIGNMENT=true, TOPOLOGY=/zone/instance, TOPOLOGY_AWARE_ENABLED=true}{REBALANCE_PREFERENCE={EVENNESS=0, LESS_MOVEMENT=10}}{}, Stat=Stat {_version=4, _creationTime=1710804015571, _modifiedTime=1710804047240, _ephemeralOwner=0} Failure Type: INVALID_CLUSTER_STATUS
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.handleDelayedRebalanceMinActiveReplica(WagedRebalancer.java:428) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.emergencyRebalance(WagedRebalancer.java:501) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.computeBestPossibleAssignment(WagedRebalancer.java:339) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.computeBestPossibleStates(WagedRebalancer.java:316) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.computeNewIdealStates(WagedRebalancer.java:248) [classes/:?]
    at org.apache.helix.controller.stages.BestPossibleStateCalcStage.computeResourceBestPossibleStateWithWagedRebalancer(BestPossibleStateCalcStage.java:406) [classes/:?]
    at org.apache.helix.controller.stages.BestPossibleStateCalcStage.compute(BestPossibleStateCalcStage.java:258) [classes/:?]
    at org.apache.helix.controller.stages.BestPossibleStateCalcStage.process(BestPossibleStateCalcStage.java:91) [classes/:?]
    at org.apache.helix.controller.pipeline.Pipeline.handle(Pipeline.java:75) [classes/:?]
    at org.apache.helix.controller.GenericHelixController.handleEvent(GenericHelixController.java:903) [classes/:?]
    at org.apache.helix.controller.GenericHelixController$ClusterEventProcessor.run(GenericHelixController.java:1554) [classes/:?]
Caused by: java.lang.NullPointerException
    at org.apache.helix.controller.rebalancer.util.DelayedRebalanceUtil.findToBeAssignedReplicasForMinActiveReplica(DelayedRebalanceUtil.java:335) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.model.ClusterModelProvider.generateClusterModel(ClusterModelProvider.java:257) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.model.ClusterModelProvider.generateClusterModelForDelayedRebalanceOverwrites(ClusterModelProvider.java:82) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.handleDelayedRebalanceMinActiveReplica(WagedRebalancer.java:415) ~[classes/:?]
    ... 10 more
94184 [main] ERROR org.apache.helix.controller.rebalancer.waged.WagedRebalancer [] - Failed to compute for delayed rebalance overwrites in cluster CLUSTER_TestWagedClusterExpansionWithAddingResourcesBeforeInstances
94188 [main] ERROR org.apache.helix.controller.rebalancer.waged.WagedRebalancer [] - Failed to calculate the new assignments.
org.apache.helix.HelixRebalanceException: Failed to compute for delayed rebalance overwrites in cluster ZnRecord=CLUSTER_TestWagedClusterExpansionWithAddingResourcesBeforeInstances, {DELAY_REBALANCE_ENABLED=true, DELAY_REBALANCE_TIME=3000000, FAULT_ZONE_TYPE=zone, PERSIST_BEST_POSSIBLE_ASSIGNMENT=true, TOPOLOGY=/zone/instance, TOPOLOGY_AWARE_ENABLED=true}{REBALANCE_PREFERENCE={EVENNESS=0, LESS_MOVEMENT=10}}{}, Stat=Stat {_version=4, _creationTime=1710804015571, _modifiedTime=1710804047240, _ephemeralOwner=0} Failure Type: INVALID_CLUSTER_STATUS
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.handleDelayedRebalanceMinActiveReplica(WagedRebalancer.java:428) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.emergencyRebalance(WagedRebalancer.java:514) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.computeBestPossibleAssignment(WagedRebalancer.java:339) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.computeBestPossibleStates(WagedRebalancer.java:316) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.computeNewIdealStates(WagedRebalancer.java:248) [classes/:?]
    at org.apache.helix.controller.stages.BestPossibleStateCalcStage.computeResourceBestPossibleStateWithWagedRebalancer(BestPossibleStateCalcStage.java:406) [classes/:?]
    at org.apache.helix.controller.stages.BestPossibleStateCalcStage.compute(BestPossibleStateCalcStage.java:258) [classes/:?]
    at org.apache.helix.controller.stages.BestPossibleStateCalcStage.process(BestPossibleStateCalcStage.java:91) [classes/:?]
    at org.apache.helix.util.RebalanceUtil.runStage(RebalanceUtil.java:235) [classes/:?]
    at org.apache.helix.tools.ClusterVerifiers.BestPossibleExternalViewVerifier.calcBestPossState(BestPossibleExternalViewVerifier.java:444) [classes/:?]
    at org.apache.helix.tools.ClusterVerifiers.BestPossibleExternalViewVerifier.verifyState(BestPossibleExternalViewVerifier.java:293) [classes/:?]
    at org.apache.helix.tools.ClusterVerifiers.ZkHelixClusterVerifier.verifyByPolling(ZkHelixClusterVerifier.java:209) [classes/:?]
    at org.apache.helix.tools.ClusterVerifiers.ZkHelixClusterVerifier.verifyByPolling(ZkHelixClusterVerifier.java:229) [classes/:?]
    at org.apache.helix.integration.rebalancer.WagedRebalancer.TestWagedClusterExpansionWithAddingResourcesBeforeInstances.testExpandClusterWithResourceWithoutInstances(TestWagedClusterExpansionWithAddingResourcesBeforeInstances.java:151) [test-classes/:?]
    at jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:?]
    at jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:?]
    at jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:?]
Caused by: java.lang.NullPointerException
    at org.apache.helix.controller.rebalancer.util.DelayedRebalanceUtil.findToBeAssignedReplicasForMinActiveReplica(DelayedRebalanceUtil.java:335) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.model.ClusterModelProvider.generateClusterModel(ClusterModelProvider.java:257) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.model.ClusterModelProvider.generateClusterModelForDelayedRebalanceOverwrites(ClusterModelProvider.java:82) ~[classes/:?]
    at org.apache.helix.controller.rebalancer.waged.WagedRebalancer.handleDelayedRebalanceMinActiveReplica(WagedRebalancer.java:415) ~[classes/:?]
    ... 37 more

Tests

[x] The following tests are written for this issue:

(List the names of added unit/integration tests)

TestWagedClusterExpansionWithAddingResourcesBeforeInstances.testExpandClusterWithResourceWithoutInstances
The following is the result of the "mvn test" command on the appropriate module:

➜  helix_os_hk git:(hkandwal/waged-adding-resources-without-capacity) mvn clean install -Dmaven.test.skip.exec=true && mvn test -o -Dtest=TestWagedClusterExpansionWithAddingResourcesBeforeInstances -pl=helix-core

[INFO] --- surefire:3.0.0-M3:test (default-test) @ helix-core ---
[INFO] 
[INFO] -------------------------------------------------------
[INFO]  T E S T S
[INFO] -------------------------------------------------------
[INFO] Running org.apache.helix.integration.rebalancer.WagedRebalancer.TestWagedClusterExpansionWithAddingResourcesBeforeInstances
Start zookeeper at localhost:2183 in thread main
AfterClass: TestWagedClusterExpansionWithAddingResourcesBeforeInstances called.
Shut down zookeeper at port 2183 in thread main
[INFO] Tests run: 1, Failures: 0, Errors: 0, Skipped: 0, Time elapsed: 17.354 s - in org.apache.helix.integration.rebalancer.WagedRebalancer.TestWagedClusterExpansionWithAddingResourcesBeforeInstances
[INFO] 
[INFO] Results:
[INFO] 
[INFO] Tests run: 1, Failures: 0, Errors: 0, Skipped: 0
[INFO] 
[INFO] 
[INFO] --- jacoco:0.8.6:report (generate-code-coverage-report) @ helix-core ---
[INFO] Loading execution data file /Users/hkandwal/Documents/workspaces/projects/helix_os_hk/helix-core/target/jacoco.exec
[INFO] Analyzed bundle 'Apache Helix :: Core' with 950 classes
[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO] ------------------------------------------------------------------------
[INFO] Total time:  54.225 s
[INFO] Finished at: 2024-03-27T16:29:31-07:00
[INFO] ------------------------------------------------------------------------

Changes that Break Backward Compatibility (Optional)

My PR contains changes that break backward compatibility or previous assumptions for certain methods or API. They include:

(Consider including all behavior changes for public methods or API. Also include these changes in merge description so that other developers are aware of these changes. This allows them to make relevant code changes in feature branches accounting for the new method/API behavior.)

Documentation (Optional)

In case of new functionality, my PR adds documentation in the following wiki page:

(Link the GitHub wiki you added)

Commits

My commits all reference appropriate Apache Helix GitHub issues in their subject lines. In addition, my commits follow the guidelines from "How to write a good git commit message":
1. Subject is separated from body by a blank line
2. Subject is limited to 50 characters (not including Jira issue reference)
3. Subject does not end with a period
4. Subject uses the imperative mood ("add", not "adding")
5. Body wraps at 72 characters
6. Body explains "what" and "why", not "how"

Code Quality

My diff has been formatted using helix-style.xml (helix-style-intellij.xml if IntelliJ IDE is used)

junkaixue commented 4 months ago

Is this just for adding a test?

himanshukandwal commented 4 months ago

Is this just for adding a test?

Hey @junkaixue yes, I first reproduced it using the test, now added the fix. Would you pls review it when get a chance.

himanshukandwal commented 4 months ago

Please remove WIP from title

Done

himanshukandwal commented 4 months ago

This PR has been reviewed and approved by @zpinto.

Final Commit Message: When adding a new WAGED resource with a tag and without any instances against that tag, we are observing NPE coming from the system. To solve this issue we are adding a check in the ResourceComputationStage to have such resources excluded from the pipeline computation and only be considered when there are actual resource partitions (>0) to be assigned to the instances.

junkaixue commented 4 months ago

lgtm. Let's wait for the tests.

himanshukandwal commented 4 months ago

Hey @junkaixue, the PR CI successful.

Also ran the full CI suite in my repo and thats successful as well: https://github.com/himanshukandwal/helix/actions/runs/8471701884/job/23212239715

apache / helix