[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=446440=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-446440 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 16/Jun/20 11:44 Start Date: 16/Jun/20 11:44 Worklog Time Spent: 10m Work Description: github-actions[bot] closed pull request #587: URL: https://github.com/apache/hive/pull/587 This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 446440) Time Spent: 2.5h (was: 2h 20m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch, HIVE-21213.04.patch, HIVE-21213.05.patch > > Time Spent: 2.5h > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=442975=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-442975 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 09/Jun/20 16:17 Start Date: 09/Jun/20 16:17 Worklog Time Spent: 10m Work Description: github-actions[bot] commented on pull request #587: URL: https://github.com/apache/hive/pull/587#issuecomment-641143683 This pull request has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Feel free to reach out on the d...@hive.apache.org list if the patch is in need of reviews. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 442975) Time Spent: 2h 20m (was: 2h 10m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch, HIVE-21213.04.patch, HIVE-21213.05.patch > > Time Spent: 2h 20m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=361580=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-361580 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 20/Dec/19 10:35 Start Date: 20/Dec/19 10:35 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r360316128 ## File path: itests/hive-unit/src/test/java/org/apache/hadoop/hive/ql/parse/TestReplicationScenariosAcidTables.java ## @@ -670,4 +678,63 @@ public void testMultiDBTxn() throws Throwable { replica.run("drop database " + dbName1 + " cascade"); replica.run("drop database " + dbName2 + " cascade"); } + + private void runCompaction(String dbName, String tblName, CompactionType compactionType) throws Throwable { +HiveConf hiveConf = new HiveConf(primary.getConf()); +TxnStore txnHandler = TxnUtils.getTxnStore(hiveConf); +txnHandler.compact(new CompactionRequest(dbName, tblName, compactionType)); +hiveConf.setBoolVar(HiveConf.ConfVars.COMPACTOR_CRUD_QUERY_BASED, false); +runWorker(hiveConf); +runCleaner(hiveConf); + } + + private FileStatus[] getDirsInTableLoc(WarehouseInstance wh, String db, String table) throws Throwable { +Path tblLoc = new Path(wh.getTable(db, table).getSd().getLocation()); +FileSystem fs = tblLoc.getFileSystem(wh.getConf()); +return fs.listStatus(tblLoc, EximUtil.getDirectoryFilter(fs)); + } + + @Test + public void testAcidTablesBootstrapWithCompaction() throws Throwable { + String tableName = testName.getMethodName(); + primary.run("use " + primaryDbName) +.run("create table " + tableName + " (id int) clustered by(id) into 3 buckets stored as orc " + +"tblproperties (\"transactional\"=\"true\")") +.run("insert into " + tableName + " values(1)") +.run("insert into " + tableName + " values(2)"); +runCompaction(primaryDbName, tableName, CompactionType.MAJOR); Review comment: Same as above. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 361580) Time Spent: 2h 10m (was: 2h) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 2h 10m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=361579=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-361579 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 20/Dec/19 10:35 Start Date: 20/Dec/19 10:35 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r360316053 ## File path: itests/hive-unit/src/test/java/org/apache/hadoop/hive/ql/parse/TestReplicationScenariosAcidTables.java ## @@ -670,4 +678,63 @@ public void testMultiDBTxn() throws Throwable { replica.run("drop database " + dbName1 + " cascade"); replica.run("drop database " + dbName2 + " cascade"); } + + private void runCompaction(String dbName, String tblName, CompactionType compactionType) throws Throwable { +HiveConf hiveConf = new HiveConf(primary.getConf()); +TxnStore txnHandler = TxnUtils.getTxnStore(hiveConf); +txnHandler.compact(new CompactionRequest(dbName, tblName, compactionType)); +hiveConf.setBoolVar(HiveConf.ConfVars.COMPACTOR_CRUD_QUERY_BASED, false); +runWorker(hiveConf); +runCleaner(hiveConf); + } + + private FileStatus[] getDirsInTableLoc(WarehouseInstance wh, String db, String table) throws Throwable { +Path tblLoc = new Path(wh.getTable(db, table).getSd().getLocation()); +FileSystem fs = tblLoc.getFileSystem(wh.getConf()); +return fs.listStatus(tblLoc, EximUtil.getDirectoryFilter(fs)); + } + + @Test + public void testAcidTablesBootstrapWithCompaction() throws Throwable { + String tableName = testName.getMethodName(); + primary.run("use " + primaryDbName) +.run("create table " + tableName + " (id int) clustered by(id) into 3 buckets stored as orc " + +"tblproperties (\"transactional\"=\"true\")") Review comment: In case the logic changes in future, we need a test to safe guard. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 361579) Time Spent: 2h (was: 1h 50m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 2h > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=361577=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-361577 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 20/Dec/19 10:34 Start Date: 20/Dec/19 10:34 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r360315778 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. + if (AcidUtils.getVisibilityTxnId(subDir) > 0) { Review comment: Thanks for the explanation. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 361577) Time Spent: 1h 40m (was: 1.5h) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 1h 40m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=361578=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-361578 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 20/Dec/19 10:35 Start Date: 20/Dec/19 10:35 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r360315864 ## File path: standalone-metastore/metastore-server/src/main/java/org/apache/hadoop/hive/metastore/txn/TxnHandler.java ## @@ -1467,7 +1467,7 @@ public void replTableWriteIdState(ReplTblWriteIdStateRequest rqst) throws MetaEx // Schedule Major compaction on all the partitions/table to clean aborted data if (numAbortedWrites > 0) { - CompactionRequest compactRqst = new CompactionRequest(rqst.getDbName(), rqst.getTableName(), + CompactionRequest compactRqst = new CompactionRequest(dbName, tblName, Review comment: Ok. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 361578) Time Spent: 1h 50m (was: 1h 40m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 1h 50m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=361576=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-361576 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 20/Dec/19 10:34 Start Date: 20/Dec/19 10:34 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r360315681 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. + if (AcidUtils.getVisibilityTxnId(subDir) > 0) { +if (subDir.startsWith(AcidUtils.BASE_PREFIX)) { + AcidUtils.ParsedBase pb = AcidUtils.ParsedBase.parseBase(new Path(subDir)); + destination = new Path(destination, AcidUtils.baseDir(pb.getWriteId())); +} else if (subDir.startsWith(AcidUtils.DELTA_PREFIX)) { + AcidUtils.ParsedDeltaLight pdl = AcidUtils.ParsedDeltaLight.parse(new Path(subDir)); + destination = new Path(destination, AcidUtils.deltaSubdir(pdl.getMinWriteId(), pdl.getMaxWriteId())); +} else if (subDir.startsWith(AcidUtils.DELETE_DELTA_PREFIX)) { + AcidUtils.ParsedDeltaLight pdl = AcidUtils.ParsedDeltaLight.parse(new Path(subDir)); + destination = new Path(destination, AcidUtils.deleteDeltaSubdir(pdl.getMinWriteId(), pdl.getMaxWriteId())); Review comment: Hmm, when the statement ids are actually used, this might become a problem. For now this looks ok. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 361576) Time Spent: 1.5h (was: 1h 20m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 1.5h > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=361574=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-361574 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 20/Dec/19 10:33 Start Date: 20/Dec/19 10:33 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r360315217 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. Review comment: The real question is whether there's a bug here, if compaction happens on the source as well as on the target. It would be better to clarify that this login needs to be worked again when incremental replication will be worked again. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 361574) Time Spent: 1h 10m (was: 1h) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 1h 10m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=361575=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-361575 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 20/Dec/19 10:33 Start Date: 20/Dec/19 10:33 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r360315411 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. + if (AcidUtils.getVisibilityTxnId(subDir) > 0) { +if (subDir.startsWith(AcidUtils.BASE_PREFIX)) { + AcidUtils.ParsedBase pb = AcidUtils.ParsedBase.parseBase(new Path(subDir)); + destination = new Path(destination, AcidUtils.baseDir(pb.getWriteId())); Review comment: Thanks for the explanation. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 361575) Time Spent: 1h 20m (was: 1h 10m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 1h 20m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354254=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354254 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354280694 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. + if (AcidUtils.getVisibilityTxnId(subDir) > 0) { Review comment: What does it mean to have visibility txn id to be less than 0? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354254) Time Spent: 40m (was: 0.5h) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 40m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354258=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354258 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354286755 ## File path: itests/hive-unit/src/test/java/org/apache/hadoop/hive/ql/parse/TestReplicationScenariosAcidTables.java ## @@ -670,4 +678,63 @@ public void testMultiDBTxn() throws Throwable { replica.run("drop database " + dbName1 + " cascade"); replica.run("drop database " + dbName2 + " cascade"); } + + private void runCompaction(String dbName, String tblName, CompactionType compactionType) throws Throwable { +HiveConf hiveConf = new HiveConf(primary.getConf()); +TxnStore txnHandler = TxnUtils.getTxnStore(hiveConf); +txnHandler.compact(new CompactionRequest(dbName, tblName, compactionType)); +hiveConf.setBoolVar(HiveConf.ConfVars.COMPACTOR_CRUD_QUERY_BASED, false); +runWorker(hiveConf); +runCleaner(hiveConf); + } + + private FileStatus[] getDirsInTableLoc(WarehouseInstance wh, String db, String table) throws Throwable { +Path tblLoc = new Path(wh.getTable(db, table).getSd().getLocation()); +FileSystem fs = tblLoc.getFileSystem(wh.getConf()); +return fs.listStatus(tblLoc, EximUtil.getDirectoryFilter(fs)); + } + + @Test + public void testAcidTablesBootstrapWithCompaction() throws Throwable { + String tableName = testName.getMethodName(); + primary.run("use " + primaryDbName) +.run("create table " + tableName + " (id int) clustered by(id) into 3 buckets stored as orc " + +"tblproperties (\"transactional\"=\"true\")") +.run("insert into " + tableName + " values(1)") +.run("insert into " + tableName + " values(2)"); +runCompaction(primaryDbName, tableName, CompactionType.MAJOR); Review comment: What about a minor compaction? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354258) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 1h > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354255=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354255 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354281539 ## File path: standalone-metastore/metastore-server/src/main/java/org/apache/hadoop/hive/metastore/txn/TxnHandler.java ## @@ -1467,7 +1467,7 @@ public void replTableWriteIdState(ReplTblWriteIdStateRequest rqst) throws MetaEx // Schedule Major compaction on all the partitions/table to clean aborted data if (numAbortedWrites > 0) { - CompactionRequest compactRqst = new CompactionRequest(rqst.getDbName(), rqst.getTableName(), + CompactionRequest compactRqst = new CompactionRequest(dbName, tblName, Review comment: Why do we need this change here? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354255) Time Spent: 50m (was: 40m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 50m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354252=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354252 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354277500 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. + if (AcidUtils.getVisibilityTxnId(subDir) > 0) { +if (subDir.startsWith(AcidUtils.BASE_PREFIX)) { + AcidUtils.ParsedBase pb = AcidUtils.ParsedBase.parseBase(new Path(subDir)); + destination = new Path(destination, AcidUtils.baseDir(pb.getWriteId())); Review comment: Looks like, we are removing the transaction id from the compactor and just leaving writeId there. Are we replicating the corresponding transaction? How do we know whether this writeId is visible. I am assuming that a failed compaction will leave a directory behind but its writeId won't be visible. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354252) Time Spent: 0.5h (was: 20m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 0.5h > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354256=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354256 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354286299 ## File path: itests/hive-unit/src/test/java/org/apache/hadoop/hive/ql/parse/TestReplicationScenariosAcidTables.java ## @@ -670,4 +678,63 @@ public void testMultiDBTxn() throws Throwable { replica.run("drop database " + dbName1 + " cascade"); replica.run("drop database " + dbName2 + " cascade"); } + + private void runCompaction(String dbName, String tblName, CompactionType compactionType) throws Throwable { +HiveConf hiveConf = new HiveConf(primary.getConf()); +TxnStore txnHandler = TxnUtils.getTxnStore(hiveConf); +txnHandler.compact(new CompactionRequest(dbName, tblName, compactionType)); +hiveConf.setBoolVar(HiveConf.ConfVars.COMPACTOR_CRUD_QUERY_BASED, false); +runWorker(hiveConf); +runCleaner(hiveConf); + } + + private FileStatus[] getDirsInTableLoc(WarehouseInstance wh, String db, String table) throws Throwable { +Path tblLoc = new Path(wh.getTable(db, table).getSd().getLocation()); +FileSystem fs = tblLoc.getFileSystem(wh.getConf()); +return fs.listStatus(tblLoc, EximUtil.getDirectoryFilter(fs)); + } + + @Test + public void testAcidTablesBootstrapWithCompaction() throws Throwable { + String tableName = testName.getMethodName(); + primary.run("use " + primaryDbName) +.run("create table " + tableName + " (id int) clustered by(id) into 3 buckets stored as orc " + +"tblproperties (\"transactional\"=\"true\")") +.run("insert into " + tableName + " values(1)") +.run("insert into " + tableName + " values(2)"); +runCompaction(primaryDbName, tableName, CompactionType.MAJOR); +WarehouseInstance.Tuple bootstrapDump = primary.dump(primaryDbName, null); +replica.load(replicatedDbName, bootstrapDump.dumpLocation); +replica.run("use " + replicatedDbName) +.run("show tables") +.verifyResults(new String[] {tableName}) +.run("repl status " + replicatedDbName) +.verifyResult(bootstrapDump.lastReplicationId) +.run("select id from " + tableName + " order by id") +.verifyResults(new String[]{"1", "2"}); + +FileStatus[] dirsInLoadPath = getDirsInTableLoc(primary, primaryDbName, tableName); +long writeId = -1; +for (FileStatus fileStatus : dirsInLoadPath) { + if (fileStatus.getPath().getName().startsWith(AcidUtils.BASE_PREFIX)) { +writeId = AcidUtils.ParsedBase.parseBase(fileStatus.getPath()).getWriteId(); + assertTrue(AcidUtils.getVisibilityTxnId(fileStatus.getPath().getName()) != -1); +break; + } +} +//compaction is done so there should be a base directory. +assertTrue(writeId != -1); + +dirsInLoadPath = getDirsInTableLoc(replica, replicatedDbName, tableName); +for (FileStatus fileStatus : dirsInLoadPath) { + if (fileStatus.getPath().getName().startsWith(AcidUtils.BASE_PREFIX)) { +assertTrue(writeId == AcidUtils.ParsedBase.parseBase(fileStatus.getPath()).getWriteId()); + assertTrue(AcidUtils.getVisibilityTxnId(fileStatus.getPath().getName()) == -1); +writeId = -1; +break; + } +} +//make sure that it has done the verification. +assertTrue(writeId == -1); Review comment: Using writeId again for verification saves a variable but it's not so much readable. May be you want to save writeId on the replica in a separate variable and compare the writeId from source that on the target. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354256) Time Spent: 1h (was: 50m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354253=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354253 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354280208 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. + if (AcidUtils.getVisibilityTxnId(subDir) > 0) { +if (subDir.startsWith(AcidUtils.BASE_PREFIX)) { + AcidUtils.ParsedBase pb = AcidUtils.ParsedBase.parseBase(new Path(subDir)); + destination = new Path(destination, AcidUtils.baseDir(pb.getWriteId())); +} else if (subDir.startsWith(AcidUtils.DELTA_PREFIX)) { + AcidUtils.ParsedDeltaLight pdl = AcidUtils.ParsedDeltaLight.parse(new Path(subDir)); + destination = new Path(destination, AcidUtils.deltaSubdir(pdl.getMinWriteId(), pdl.getMaxWriteId())); +} else if (subDir.startsWith(AcidUtils.DELETE_DELTA_PREFIX)) { + AcidUtils.ParsedDeltaLight pdl = AcidUtils.ParsedDeltaLight.parse(new Path(subDir)); + destination = new Path(destination, AcidUtils.deleteDeltaSubdir(pdl.getMinWriteId(), pdl.getMaxWriteId())); Review comment: We are ignoring the statement id, I think because it's -1 always. Is that future proof? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354253) Time Spent: 40m (was: 0.5h) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 40m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354251=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354251 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354223809 ## File path: ql/src/java/org/apache/hadoop/hive/ql/parse/repl/CopyUtils.java ## @@ -463,7 +464,29 @@ public static Path getCopyDestination(ReplChangeManager.FileInfo fileInfo, Path String[] subDirs = fileInfo.getSubDir().split(Path.SEPARATOR); Path destination = destRoot; for (String subDir: subDirs) { - destination = new Path(destination, subDir); + // If the directory is created by compactor, then the directory will have the transaction id also. + // In case of replication, the same txn id can not be used at target, as the txn with same id might be a + // aborted or live txn at target. + // In case of bootstrap load, we copy only the committed data, so the directory with only write id + // can be created. The validity txn id can be removed from the directory name. + // TODO : Support for incremental load flow. This can be done once replication of compaction is decided. Review comment: Is this TODO resolved? If yes, please remove it. If it needs to be resolved, may be we should resolve it. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354251) Time Spent: 20m (was: 10m) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 20m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=354257=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-354257 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 05/Dec/19 12:36 Start Date: 05/Dec/19 12:36 Worklog Time Spent: 10m Work Description: ashutosh-bapat commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587#discussion_r354286550 ## File path: itests/hive-unit/src/test/java/org/apache/hadoop/hive/ql/parse/TestReplicationScenariosAcidTables.java ## @@ -670,4 +678,63 @@ public void testMultiDBTxn() throws Throwable { replica.run("drop database " + dbName1 + " cascade"); replica.run("drop database " + dbName2 + " cascade"); } + + private void runCompaction(String dbName, String tblName, CompactionType compactionType) throws Throwable { +HiveConf hiveConf = new HiveConf(primary.getConf()); +TxnStore txnHandler = TxnUtils.getTxnStore(hiveConf); +txnHandler.compact(new CompactionRequest(dbName, tblName, compactionType)); +hiveConf.setBoolVar(HiveConf.ConfVars.COMPACTOR_CRUD_QUERY_BASED, false); +runWorker(hiveConf); +runCleaner(hiveConf); + } + + private FileStatus[] getDirsInTableLoc(WarehouseInstance wh, String db, String table) throws Throwable { +Path tblLoc = new Path(wh.getTable(db, table).getSd().getLocation()); +FileSystem fs = tblLoc.getFileSystem(wh.getConf()); +return fs.listStatus(tblLoc, EximUtil.getDirectoryFilter(fs)); + } + + @Test + public void testAcidTablesBootstrapWithCompaction() throws Throwable { + String tableName = testName.getMethodName(); + primary.run("use " + primaryDbName) +.run("create table " + tableName + " (id int) clustered by(id) into 3 buckets stored as orc " + +"tblproperties (\"transactional\"=\"true\")") Review comment: Should we add a test for partitioned table as well? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 354257) > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Bug > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch, HIVE-21213.02.patch, > HIVE-21213.03.patch > > Time Spent: 1h > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Work logged] (HIVE-21213) Acid table bootstrap replication needs to handle directory created by compaction with txn id
[ https://issues.apache.org/jira/browse/HIVE-21213?focusedWorklogId=221288=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-221288 ] ASF GitHub Bot logged work on HIVE-21213: - Author: ASF GitHub Bot Created on: 01/Apr/19 13:29 Start Date: 01/Apr/19 13:29 Worklog Time Spent: 10m Work Description: maheshk114 commented on pull request #587: HIVE-21213 : Acid table bootstrap replication needs to handle directory created by compaction with txn id URL: https://github.com/apache/hive/pull/587 … This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org Issue Time Tracking --- Worklog Id: (was: 221288) Time Spent: 10m Remaining Estimate: 0h > Acid table bootstrap replication needs to handle directory created by > compaction with txn id > > > Key: HIVE-21213 > URL: https://issues.apache.org/jira/browse/HIVE-21213 > Project: Hive > Issue Type: Sub-task > Components: Hive, HiveServer2, repl >Reporter: mahesh kumar behera >Assignee: mahesh kumar behera >Priority: Major > Labels: pull-request-available > Attachments: HIVE-21213.01.patch > > Time Spent: 10m > Remaining Estimate: 0h > > The current implementation of compaction uses the txn id in the directory > name. This is used to isolate the queries from reading the directory until > compaction has finished and to avoid the compactor marking used earlier. In > case of replication, during bootstrap , directory is copied as it is with the > same name from source to destination cluster. But the directory created by > compaction with txn id can not be copied as the txn list at target may be > different from source. The txn id which is valid at source may be an aborted > txn at target. So conversion logic is required to create a new directory with > valid txn at target and dump the data to the newly created directory. -- This message was sent by Atlassian JIRA (v7.6.3#76005)