[jira] [Comment Edited] (HUDI-690) filtercompletedInstants in HudiSnapshotCopier not working as expected for MOR tables
[
https://issues.apache.org/jira/browse/HUDI-690?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17116251#comment-17116251
]
Raymond Xu edited comment on HUDI-690 at 5/25/20, 8:41 PM:
---
It is the `getCommitsTimeline()` that does not include compaction instants.
Changing to`getCommitsAndCompactionTimeline()` should do the job.
[~jomeke] would you be able to verify the change with the problematic dataset
that you had? You may simply apply the 2-line change to
hudi-utilities/src/main/java/org/apache/hudi/utilities/HoodieSnapshotExporter.java
as shown in
[https://github.com/apache/hudi/pull/1667/files#diff-5ba8b968112135426aa10a7660d9b248]
was (Author: rxu):
It is the `getCommitsTimeline()` that does not include compaction instants.
Changing to`getCommitsAndCompactionTimeline()` should do the job.
[~jomeke] would you be able to verify the change with the problematic dataset
that you had? You may simply apply the 2-line change to
`HudiSnapshotCopier.java` as shown in
[https://github.com/apache/hudi/pull/1667/files#diff-5ba8b968112135426aa10a7660d9b248]
> filtercompletedInstants in HudiSnapshotCopier not working as expected for MOR
> tables
>
>
> Key: HUDI-690
> URL: https://issues.apache.org/jira/browse/HUDI-690
> Project: Apache Hudi
> Issue Type: Improvement
>Reporter: Jasmine Omeke
>Assignee: Raymond Xu
>Priority: Major
> Labels: bug-bash-0.6.0
> Fix For: 0.6.0
>
>
> Hi. I encountered an error while using the HudiSnapshotCopier class to make a
> Backup of merge on read tables:
> [https://github.com/apache/incubator-hudi/blob/release-0.5.0/hudi-utilities/src/main/java/org/apache/hudi/utilities/HoodieSnapshotCopier.java]
>
> The error:
>
> {code:java}
> 20/03/09 15:43:19 INFO AmazonHttpClient: Configuring Proxy. Proxy Host:
> web-proxy.bt.local Proxy Port: 3128
> 20/03/09 15:43:19 INFO HoodieTableConfig: Loading dataset properties from
> /.hoodie/hoodie.properties
> 20/03/09 15:43:19 INFO AmazonHttpClient: Configuring Proxy. Proxy Host:
> web-proxy.bt.local Proxy Port: 3128
> 20/03/09 15:43:19 INFO HoodieTableMetaClient: Finished Loading Table of type
> MERGE_ON_READ from
> 20/03/09 15:43:20 INFO HoodieActiveTimeline: Loaded instants
> java.util.stream.ReferencePipeline$Head@77f7352a
> 20/03/09 15:43:21 INFO YarnSchedulerBackend$YarnDriverEndpoint: Registered
> executor NettyRpcEndpointRef(spark-client://Executor) (10.49.26.74:40894)
> with ID 2
> 20/03/09 15:43:21 INFO ExecutorAllocationManager: New executor 2 has
> registered (new total is 1)
> 20/03/09 15:43:21 INFO BlockManagerMasterEndpoint: Registering block manager
> ip-10-49-26-74.us-east-2.compute.internal:32831 with 12.4 GB RAM,
> BlockManagerId(2, ip-10-49-26-74.us-east-2.compute.internal, 3283
> 1, None)
> 20/03/09 15:43:21 INFO YarnSchedulerBackend$YarnDriverEndpoint: Registered
> executor NettyRpcEndpointRef(spark-client://Executor) (10.49.26.74:40902)
> with ID 4
> 20/03/09 15:43:21 INFO ExecutorAllocationManager: New executor 4 has
> registered (new total is 2)Exception in thread "main"
> java.lang.IllegalStateException: Hudi File Id
> (HoodieFileGroupId{partitionPath='created_at_month=2020-03',
> fileId='7104bb0b-20f6-4dec-981b-c11bf20ade4a-0'}) has more than 1 pending
> compactions. Instants: (20200309011643,{"baseInstantTime": "20200308213934",
> "deltaFilePaths":
> [".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.1_3-751289-170568496",
> ".7104bb0b-20f6-4dec-981b-c11
> bf20ade4a-0_20200308213934.log.2_3-761601-172985464",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.3_1-772174-175483657",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.4_2-782377-
> 177872977",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.5_1-790994-179909226"],
> "dataFilePath":
> "7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_0-746201-169642460_20200308213934.parquet",
> "fileId": "7
> 104bb0b-20f6-4dec-981b-c11bf20ade4a-0", "partitionPath":
> "created_at_month=2020-03", "metrics": {"TOTAL_LOG_FILES": 5.0,
> "TOTAL_IO_READ_MB": 512.0, "TOTAL_LOG_FILES_SIZE": 33789.0,
> "TOTAL_IO_WRITE_MB": 512.0,
> "TOTAL_IO_MB": 1024.0, "TOTAL_LOG_FILE_SIZE": 33789.0}}),
> (20200308213934,{"baseInstantTime": "20200308180755", "deltaFilePaths":
> [".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.1_3-696047-158157865",
>
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.2_2-706457-160605423",
>
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.3_1-716977-163056814",
> ".7104bb0b-20f6-4dec-981b-c11bf20ad
> e4a-0_20200308180755.log.4_3-727192-165430450",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.5_3-737755-167913339"],
> "dataFilePath":
[jira] [Comment Edited] (HUDI-690) filtercompletedInstants in HudiSnapshotCopier not working as expected for MOR tables
[
https://issues.apache.org/jira/browse/HUDI-690?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17116251#comment-17116251
]
Raymond Xu edited comment on HUDI-690 at 5/25/20, 8:39 PM:
---
It is the `getCommitsTimeline()` that does not include compaction instants.
Changing to`getCommitsAndCompactionTimeline()` should do the job.
[~jomeke] would you be able to verify the change with the problematic dataset
that you had? You may simply apply the 2-line change to
`HudiSnapshotCopier.java` as shown in
[https://github.com/apache/hudi/pull/1667/files#diff-5ba8b968112135426aa10a7660d9b248]
was (Author: rxu):
It is the `getCommitsTimeline()` that does not include compaction instants.
Changing to`getCommitsAndCompactionTimeline()` should do the job.
[~jomeke] would you be able to verify the change with the problematic dataset
that you had? You may simply apply the changes as shown in
[https://github.com/apache/hudi/pull/1667/files#diff-5ba8b968112135426aa10a7660d9b248]
> filtercompletedInstants in HudiSnapshotCopier not working as expected for MOR
> tables
>
>
> Key: HUDI-690
> URL: https://issues.apache.org/jira/browse/HUDI-690
> Project: Apache Hudi
> Issue Type: Improvement
>Reporter: Jasmine Omeke
>Assignee: Raymond Xu
>Priority: Major
> Labels: bug-bash-0.6.0
> Fix For: 0.6.0
>
>
> Hi. I encountered an error while using the HudiSnapshotCopier class to make a
> Backup of merge on read tables:
> [https://github.com/apache/incubator-hudi/blob/release-0.5.0/hudi-utilities/src/main/java/org/apache/hudi/utilities/HoodieSnapshotCopier.java]
>
> The error:
>
> {code:java}
> 20/03/09 15:43:19 INFO AmazonHttpClient: Configuring Proxy. Proxy Host:
> web-proxy.bt.local Proxy Port: 3128
> 20/03/09 15:43:19 INFO HoodieTableConfig: Loading dataset properties from
> /.hoodie/hoodie.properties
> 20/03/09 15:43:19 INFO AmazonHttpClient: Configuring Proxy. Proxy Host:
> web-proxy.bt.local Proxy Port: 3128
> 20/03/09 15:43:19 INFO HoodieTableMetaClient: Finished Loading Table of type
> MERGE_ON_READ from
> 20/03/09 15:43:20 INFO HoodieActiveTimeline: Loaded instants
> java.util.stream.ReferencePipeline$Head@77f7352a
> 20/03/09 15:43:21 INFO YarnSchedulerBackend$YarnDriverEndpoint: Registered
> executor NettyRpcEndpointRef(spark-client://Executor) (10.49.26.74:40894)
> with ID 2
> 20/03/09 15:43:21 INFO ExecutorAllocationManager: New executor 2 has
> registered (new total is 1)
> 20/03/09 15:43:21 INFO BlockManagerMasterEndpoint: Registering block manager
> ip-10-49-26-74.us-east-2.compute.internal:32831 with 12.4 GB RAM,
> BlockManagerId(2, ip-10-49-26-74.us-east-2.compute.internal, 3283
> 1, None)
> 20/03/09 15:43:21 INFO YarnSchedulerBackend$YarnDriverEndpoint: Registered
> executor NettyRpcEndpointRef(spark-client://Executor) (10.49.26.74:40902)
> with ID 4
> 20/03/09 15:43:21 INFO ExecutorAllocationManager: New executor 4 has
> registered (new total is 2)Exception in thread "main"
> java.lang.IllegalStateException: Hudi File Id
> (HoodieFileGroupId{partitionPath='created_at_month=2020-03',
> fileId='7104bb0b-20f6-4dec-981b-c11bf20ade4a-0'}) has more than 1 pending
> compactions. Instants: (20200309011643,{"baseInstantTime": "20200308213934",
> "deltaFilePaths":
> [".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.1_3-751289-170568496",
> ".7104bb0b-20f6-4dec-981b-c11
> bf20ade4a-0_20200308213934.log.2_3-761601-172985464",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.3_1-772174-175483657",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.4_2-782377-
> 177872977",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.5_1-790994-179909226"],
> "dataFilePath":
> "7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_0-746201-169642460_20200308213934.parquet",
> "fileId": "7
> 104bb0b-20f6-4dec-981b-c11bf20ade4a-0", "partitionPath":
> "created_at_month=2020-03", "metrics": {"TOTAL_LOG_FILES": 5.0,
> "TOTAL_IO_READ_MB": 512.0, "TOTAL_LOG_FILES_SIZE": 33789.0,
> "TOTAL_IO_WRITE_MB": 512.0,
> "TOTAL_IO_MB": 1024.0, "TOTAL_LOG_FILE_SIZE": 33789.0}}),
> (20200308213934,{"baseInstantTime": "20200308180755", "deltaFilePaths":
> [".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.1_3-696047-158157865",
>
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.2_2-706457-160605423",
>
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.3_1-716977-163056814",
> ".7104bb0b-20f6-4dec-981b-c11bf20ad
> e4a-0_20200308180755.log.4_3-727192-165430450",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.5_3-737755-167913339"],
> "dataFilePath": "7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_0-690668-157158597_2
> 0200308180755.parquet", "fileId":
[jira] [Comment Edited] (HUDI-690) filtercompletedInstants in HudiSnapshotCopier not working as expected for MOR tables
[
https://issues.apache.org/jira/browse/HUDI-690?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17056437#comment-17056437
]
Jasmine Omeke edited comment on HUDI-690 at 3/10/20, 9:42 PM:
--
pinging to triage
[~vbalaji]
was (Author: jomeke):
[~vbalaji]
> filtercompletedInstants in HudiSnapshotCopier not working as expected for MOR
> tables
>
>
> Key: HUDI-690
> URL: https://issues.apache.org/jira/browse/HUDI-690
> Project: Apache Hudi (incubating)
> Issue Type: Improvement
>Reporter: Jasmine Omeke
>Priority: Major
>
> Hi. I encountered an error while using the HudiSnapshotCopier class to make a
> Backup of merge on read tables:
> [https://github.com/apache/incubator-hudi/blob/release-0.5.0/hudi-utilities/src/main/java/org/apache/hudi/utilities/HoodieSnapshotCopier.java]
>
> The error:
>
> {code:java}
> 20/03/09 15:43:19 INFO AmazonHttpClient: Configuring Proxy. Proxy Host:
> web-proxy.bt.local Proxy Port: 3128
> 20/03/09 15:43:19 INFO HoodieTableConfig: Loading dataset properties from
> /.hoodie/hoodie.properties
> 20/03/09 15:43:19 INFO AmazonHttpClient: Configuring Proxy. Proxy Host:
> web-proxy.bt.local Proxy Port: 3128
> 20/03/09 15:43:19 INFO HoodieTableMetaClient: Finished Loading Table of type
> MERGE_ON_READ from
> 20/03/09 15:43:20 INFO HoodieActiveTimeline: Loaded instants
> java.util.stream.ReferencePipeline$Head@77f7352a
> 20/03/09 15:43:21 INFO YarnSchedulerBackend$YarnDriverEndpoint: Registered
> executor NettyRpcEndpointRef(spark-client://Executor) (10.49.26.74:40894)
> with ID 2
> 20/03/09 15:43:21 INFO ExecutorAllocationManager: New executor 2 has
> registered (new total is 1)
> 20/03/09 15:43:21 INFO BlockManagerMasterEndpoint: Registering block manager
> ip-10-49-26-74.us-east-2.compute.internal:32831 with 12.4 GB RAM,
> BlockManagerId(2, ip-10-49-26-74.us-east-2.compute.internal, 3283
> 1, None)
> 20/03/09 15:43:21 INFO YarnSchedulerBackend$YarnDriverEndpoint: Registered
> executor NettyRpcEndpointRef(spark-client://Executor) (10.49.26.74:40902)
> with ID 4
> 20/03/09 15:43:21 INFO ExecutorAllocationManager: New executor 4 has
> registered (new total is 2)Exception in thread "main"
> java.lang.IllegalStateException: Hudi File Id
> (HoodieFileGroupId{partitionPath='created_at_month=2020-03',
> fileId='7104bb0b-20f6-4dec-981b-c11bf20ade4a-0'}) has more than 1 pending
> compactions. Instants: (20200309011643,{"baseInstantTime": "20200308213934",
> "deltaFilePaths":
> [".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.1_3-751289-170568496",
> ".7104bb0b-20f6-4dec-981b-c11
> bf20ade4a-0_20200308213934.log.2_3-761601-172985464",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.3_1-772174-175483657",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.4_2-782377-
> 177872977",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308213934.log.5_1-790994-179909226"],
> "dataFilePath":
> "7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_0-746201-169642460_20200308213934.parquet",
> "fileId": "7
> 104bb0b-20f6-4dec-981b-c11bf20ade4a-0", "partitionPath":
> "created_at_month=2020-03", "metrics": {"TOTAL_LOG_FILES": 5.0,
> "TOTAL_IO_READ_MB": 512.0, "TOTAL_LOG_FILES_SIZE": 33789.0,
> "TOTAL_IO_WRITE_MB": 512.0,
> "TOTAL_IO_MB": 1024.0, "TOTAL_LOG_FILE_SIZE": 33789.0}}),
> (20200308213934,{"baseInstantTime": "20200308180755", "deltaFilePaths":
> [".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.1_3-696047-158157865",
>
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.2_2-706457-160605423",
>
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.3_1-716977-163056814",
> ".7104bb0b-20f6-4dec-981b-c11bf20ad
> e4a-0_20200308180755.log.4_3-727192-165430450",
> ".7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_20200308180755.log.5_3-737755-167913339"],
> "dataFilePath": "7104bb0b-20f6-4dec-981b-c11bf20ade4a-0_0-690668-157158597_2
> 0200308180755.parquet", "fileId": "7104bb0b-20f6-4dec-981b-c11bf20ade4a-0",
> "partitionPath": "created_at_month=2020-03", "metrics": {"TOTAL_LOG_FILES":
> 5.0, "TOTAL_IO_READ_MB": 512.0, "TOTAL_LOG_FILES_SIZE":
> 44197.0, "TOTAL_IO_WRITE_MB": 511.0, "TOTAL_IO_MB": 1023.0,
> "TOTAL_LOG_FILE_SIZE": 44197.0}})at
> org.apache.hudi.common.util.CompactionUtils.lambda$getAllPendingCompactionOperations$5(CompactionUtils.java:161)
> at
> java.util.stream.ForEachOps$ForEachOp$OfRef.accept(ForEachOps.java:183)
> at
> java.util.stream.ForEachOps$ForEachOp$OfRef.accept(ForEachOps.java:183)
> at
> java.util.stream.ReferencePipeline$3$1.accept(ReferencePipeline.java:193)
> at java.util.Iterator.forEachRemaining(Iterator.java:116)
> at
>
