Skip to content
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.

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

[Bug] Multiple Hoodie writer commit conflict #34

Open
zymap opened this issue May 19, 2022 · 0 comments
Open

[Bug] Multiple Hoodie writer commit conflict #34

zymap opened this issue May 19, 2022 · 0 comments
Assignees

Comments

@zymap
Copy link
Member

zymap commented May 19, 2022

Describe the bug
When we run the hudi sink with multiple instances, then using it sink a partitioned topic with failover subscription mode. Both instances will consume the message from the topic. Hudi supports the concurrency mode to support multiple writers.
When we enable this feature, the hudi writer will throw fileAlreadyExists exception and failed the commit.

org.apache.hudi.exception.HoodieIOException: Failed to create file file:/tmp/integration/hudi/.hoodie/20220517094051766.commit
	at org.apache.hudi.common.table.timeline.HoodieActiveTimeline.createImmutableFileInPath(HoodieActiveTimeline.java:745) ~[hudi-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.common.table.timeline.HoodieActiveTimeline.transitionState(HoodieActiveTimeline.java:560) ~[hudi-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.common.table.timeline.HoodieActiveTimeline.transitionState(HoodieActiveTimeline.java:536) ~[hudi-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.common.table.timeline.HoodieActiveTimeline.saveAsComplete(HoodieActiveTimeline.java:183) ~[hudi-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.client.BaseHoodieWriteClient.commit(BaseHoodieWriteClient.java:270) ~[hudi-client-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.client.BaseHoodieWriteClient.commitStats(BaseHoodieWriteClient.java:234) ~[hudi-client-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.client.HoodieJavaWriteClient.commit(HoodieJavaWriteClient.java:88) ~[hudi-java-client-0.11.0.jar:0.11.0]
	at org.apache.hudi.client.HoodieJavaWriteClient.commit(HoodieJavaWriteClient.java:51) ~[hudi-java-client-0.11.0.jar:0.11.0]
	at org.apache.hudi.client.BaseHoodieWriteClient.commit(BaseHoodieWriteClient.java:206) ~[hudi-client-common-0.11.0.jar:0.11.0]
	at org.apache.pulsar.ecosystem.io.sink.hudi.BufferedConnectWriter.flushRecords(BufferedConnectWriter.java:82) ~[PqY5lYEJSWPWMDq7E5HC2Q/:?]
	at org.apache.pulsar.ecosystem.io.sink.hudi.HoodieWriter.flush(HoodieWriter.java:85) ~[PqY5lYEJSWPWMDq7E5HC2Q/:?]
	at org.apache.pulsar.ecosystem.io.sink.SinkWriter.commitIfNeed(SinkWriter.java:128) ~[PqY5lYEJSWPWMDq7E5HC2Q/:?]
	at org.apache.pulsar.ecosystem.io.sink.SinkWriter.run(SinkWriter.java:113) [PqY5lYEJSWPWMDq7E5HC2Q/:?]
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) [?:1.8.0_201]
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) [?:1.8.0_201]
	at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30) [netty-common-4.1.77.Final.jar:4.1.77.Final]
	at java.lang.Thread.run(Thread.java:748) [?:1.8.0_201]
Caused by: org.apache.hadoop.fs.FileAlreadyExistsException: File already exists: file:/tmp/integration/hudi/.hoodie/20220517094051766.commit
	at org.apache.hadoop.fs.RawLocalFileSystem.create(RawLocalFileSystem.java:315) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hadoop.fs.RawLocalFileSystem.create(RawLocalFileSystem.java:353) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hadoop.fs.ChecksumFileSystem$ChecksumFSOutputSummer.<init>(ChecksumFileSystem.java:403) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hadoop.fs.ChecksumFileSystem.create(ChecksumFileSystem.java:466) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hadoop.fs.ChecksumFileSystem.create(ChecksumFileSystem.java:445) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hadoop.fs.FileSystem.create(FileSystem.java:1125) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hadoop.fs.FileSystem.create(FileSystem.java:1105) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hadoop.fs.FileSystem.create(FileSystem.java:994) ~[hadoop-common-3.2.2.jar:?]
	at org.apache.hudi.common.fs.HoodieWrapperFileSystem.lambda$create$2(HoodieWrapperFileSystem.java:222) ~[hudi-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.common.fs.HoodieWrapperFileSystem.executeFuncWithTimeMetrics(HoodieWrapperFileSystem.java:101) ~[hudi-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.common.fs.HoodieWrapperFileSystem.create(HoodieWrapperFileSystem.java:221) ~[hudi-common-0.11.0.jar:0.11.0]
	at org.apache.hudi.common.table.timeline.HoodieActiveTimeline.createImmutableFileInPath(HoodieActiveTimeline.java:740) ~[hudi-common-0.11.0.jar:0.11.0]
	... 16 more

To Reproduce
Steps to reproduce the behavior:

  1. Create a partitioned topic pio
  2. submit a hudi writer with following configuration:
    {
  "tenant": "public",
  "namespace": "default",
  "name": "lakehouse1",
  "topicName": "pio",
  "parallelism": 2,
  "sourceSubscriptionName": "sub",
  "processingGuarantees": "EFFECTIVELY_ONCE",
  "subscriptionType": "Failover",
  "type": "hudi",
  "inputs": [
    "pio"
  ],
  "archive": "/Volumes/work/github.com/streamnative/pulsar-io-lakehouse/target/pulsar-io-lakehouse-2.9.2.0-SNAPSHOT.nar",
  "className": "org.apache.pulsar.ecosystem.io.SinkConnector",
  "configs":
  {
    "type": "hudi",
    "hoodie.table.name": "hudi-connector-test",
    "hoodie.table.type": "COPY_ON_WRITE",
    "hoodie.base.path": "file:///tmp/integration/hudi",
    "hoodie.clean.async": "true",
    "hoodie.write.concurrency.mode": "optimistic_concurrency_control",
    "hoodie.cleaner.policy.failed.writes": "LAZY",
    "hoodie.write.lock.provider": "org.apache.hudi.client.transaction.lock.ZookeeperBasedLockProvider",
    "hoodie.write.lock.zookeeper.url": "localhost",
    "hoodie.write.lock.zookeeper.port": "2181",
    "hoodie.write.lock.zookeeper.lock_key": "pulsar_hudi",
    "hoodie.write.lock.zookeeper.base_path": "/hudi",
    "hoodie.datasource.write.recordkey.field": "id",
    "hoodie.datasource.write.partitionpath.field": "id",
    "maxRecordsPerCommit": 10
  }
}
  1. produce message to the pio
  2. See error

Expected behavior
Both the commit should success

Screenshots
If applicable, add screenshots to help explain your problem.

Environment

  • OS: [e.g. Ubuntu]
  • Pulsar version: [e.g. 2.7.0]
  • Deployment: [e.g. standalone]
  • Connector/offloader/protocol handler/... version: [e.g. 2.7.0]

Additional context
Add any other context about the problem here.

@zymap zymap self-assigned this May 19, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant