Shawn Wang created KYLIN-3035: --------------------------------- Summary: How to use Kylin on EMR with S3 as hbase storage Key: KYLIN-3035 URL: https://issues.apache.org/jira/browse/KYLIN-3035 Project: Kylin Issue Type: Bug Components: Metadata Affects Versions: v2.2.0 Environment: EMR 5.5.0, Kylin 2.2.0 Reporter: Shawn Wang Assignee: Shaofeng SHI
Can somebody give an example of how to use kylin on EMR with S3 as hbase storage, which support reuse the previously built cube on new EMR after the original EMR has been terminated. My purpose is simple: 1. use transient EMR cluster to build cubes 2. use a persistent cluster to handle query requests Of course, the culsters should share same hbase storage, so I setup the cluster to use S3 as hbase storage, after 2.2.0 fix the "HFile not written to S3" issue, I have been built a sample cube successfully, using configurations: EMR: {noformat} [ { "Classification": "hbase-site", "Properties": { "hbase.rootdir": "s3://kylin-emrfs/hbase-production" } }, { "Classification": "hbase", "Properties": { "hbase.emr.storageMode": "s3" } }, { "Classification": "emrfs-site", "Properties": { "fs.s3.consistent": "true", "fs.s3.consistent.metadata.tableName": "KylinEmrFSMetadata" } } ] {noformat} kylin.propertities: {noformat} kylin.env.hdfs-working-dir=s3://kylin/kylin-emrfs/kylin-working-dir kylin.server.mode=query {noformat} Then I create a new cluster with same EMR configuration and query mode for kylin, kylin just can't startup with errors: {noformat} 2017-11-13 07:33:44,415 INFO [main-SendThread(ip-172-31-1-10.cn-north-1.compute.internal:2181)] zookeeper.ClientCnxn:876 : Socket connection established to ip-172-31-1-10.cn-north-1.compute.internal/172.31.1.10:2181, initiating session 2017-11-13 07:33:44,422 INFO [main-SendThread(ip-172-31-1-10.cn-north-1.compute.internal:2181)] zookeeper.ClientCnxn:1299 : Session establishment complete on server ip-172-31-1-10.cn-north-1.compute.internal/172.31.1.10:2181, sessionid = 0x15fb4173c100156, negotiated timeout = 40000 2017-11-13 07:33:48,380 DEBUG [main] hbase.HBaseConnection:279 : HTable 'kylin_metadata' already exists Exception in thread "main" java.lang.IllegalArgumentException: Failed to find metadata store by url: kylin_metadata@hbase at org.apache.kylin.common.persistence.ResourceStore.createResourceStore(ResourceStore.java:89) at org.apache.kylin.common.persistence.ResourceStore.getStore(ResourceStore.java:101) at org.apache.kylin.rest.service.AclTableMigrationTool.checkIfNeedMigrate(AclTableMigrationTool.java:94) at org.apache.kylin.tool.AclTableMigrationCLI.main(AclTableMigrationCLI.java:41) Caused by: org.apache.hadoop.hbase.client.RetriesExhaustedException: Failed after attempts=1, exceptions: Mon Nov 13 07:33:48 UTC 2017, RpcRetryingCaller{globalStartTime=1510558428667, pause=100, retries=1}, java.net.ConnectException: 拒绝连接 at org.apache.hadoop.hbase.client.RpcRetryingCaller.callWithRetries(RpcRetryingCaller.java:159) at org.apache.hadoop.hbase.client.HTable.get(HTable.java:864) at org.apache.hadoop.hbase.client.HTable.get(HTable.java:830) at org.apache.kylin.storage.hbase.HBaseResourceStore.internalGetFromHTable(HBaseResourceStore.java:385) at org.apache.kylin.storage.hbase.HBaseResourceStore.getFromHTable(HBaseResourceStore.java:363) at org.apache.kylin.storage.hbase.HBaseResourceStore.existsImpl(HBaseResourceStore.java:116) at org.apache.kylin.common.persistence.ResourceStore.exists(ResourceStore.java:144) at org.apache.kylin.common.persistence.ResourceStore.createResourceStore(ResourceStore.java:84) ... 3 more Caused by: java.net.ConnectException: 拒绝连接 at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method) at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:717) at org.apache.hadoop.net.SocketIOWithTimeout.connect(SocketIOWithTimeout.java:206) at org.apache.hadoop.net.NetUtils.connect(NetUtils.java:531) at org.apache.hadoop.net.NetUtils.connect(NetUtils.java:495) at org.apache.hadoop.hbase.ipc.RpcClientImpl$Connection.setupConnection(RpcClientImpl.java:416) at org.apache.hadoop.hbase.ipc.RpcClientImpl$Connection.setupIOstreams(RpcClientImpl.java:722) at org.apache.hadoop.hbase.ipc.RpcClientImpl$Connection.writeRequest(RpcClientImpl.java:909) at org.apache.hadoop.hbase.ipc.RpcClientImpl$Connection.tracedWriteRequest(RpcClientImpl.java:873) at org.apache.hadoop.hbase.ipc.RpcClientImpl.call(RpcClientImpl.java:1244) at org.apache.hadoop.hbase.ipc.AbstractRpcClient.callBlockingMethod(AbstractRpcClient.java:227) at org.apache.hadoop.hbase.ipc.AbstractRpcClient$BlockingRpcChannelImplementation.callBlockingMethod(AbstractRpcClient.java:336) at org.apache.hadoop.hbase.protobuf.generated.ClientProtos$ClientService$BlockingStub.get(ClientProtos.java:35372) at org.apache.hadoop.hbase.client.HTable$3.call(HTable.java:856) at org.apache.hadoop.hbase.client.HTable$3.call(HTable.java:847) at org.apache.hadoop.hbase.client.RpcRetryingCaller.callWithRetries(RpcRetryingCaller.java:137) ... 10 more 2017-11-13 07:33:48,709 INFO [Thread-1] client.ConnectionManager$HConnectionImplementation:2180 : Closing master protocol: MasterService 2017-11-13 07:33:48,710 INFO [Thread-1] client.ConnectionManager$HConnectionImplementation:1718 : Closing zookeeper sessionid=0x15fb4173c100156 2017-11-13 07:33:48,712 INFO [Thread-1] zookeeper.ZooKeeper:684 : Session: 0x15fb4173c100156 closed 2017-11-13 07:33:48,712 INFO [main-EventThread] zookeeper.ClientCnxn:519 : EventThread shut down for session: 0x15fb4173c100156 ERROR: Unknown error. Please check full log. {noformat} And if I change the Kylin server mode to all, Kylin can startup, but the page on port 7070 can not be opened, with similar errors as above. I am wondering If there are some other configuration I have missed, or I am just in the wrong way. It'll be many thanks if someone can give an complete example to show how to get this work done! -- This message was sent by Atlassian JIRA (v6.4.14#64029)