You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@kylin.apache.org by "suheng.cloud (JIRA)" <ji...@apache.org> on 2017/05/26 05:57:04 UTC
[jira] [Created] (KYLIN-2648) Encounter cube merge error when
deploy kylin on stand alone hbase cluster
suheng.cloud created KYLIN-2648:
-----------------------------------
Summary: Encounter cube merge error when deploy kylin on stand alone hbase cluster
Key: KYLIN-2648
URL: https://issues.apache.org/jira/browse/KYLIN-2648
Project: Kylin
Issue Type: Bug
Components: Job Engine
Affects Versions: v2.0.0
Environment: hadoop :cdh5.4.0 (both main and hbase env)
hbase : hbase-1.2.0-cdh5.7.6
hive: apache-hive-2.1.1
kylin version: 2.0
Reporter: suheng.cloud
Assignee: Dong Li
I try to deploy kylin on one node of a stand alone hbase cluster(hdfs://cdh5-mini/) which seperate from main hive cluster(hdfs://cdh5/),
According to the blog "Deploy Apache Kylin with Standalone HBase Cluster" : make sure the configurations of hadoop and hive points to main cluster,
I clone hadoop dir to another path and modify "fs.defaultFS" in core-site.xml to "hdfs://cdh5/" , and in head of kylin.sh, I export HADOOP_HOME to this new path.
So all goes well (include cube build/refresh) until I execute cube merge.
The merge error occurs at step "#9 Step Name: Garbage Collection on HDFS".
The stacktrace as follows:
2017-05-25 17:28:07,070 INFO [pool-9-thread-1] threadpool.DefaultScheduler:114 : CubingJob{id=c6709f0b-8858-4e66-a4c2-320ebc70a2e3, name=kylin_sales_cube - 20120101000000_20140201000000 - MERGE - GMT+08:00 2017-05-25 16:51:30, state=READY} prepare to schedule
2017-05-25 17:28:07,073 INFO [pool-9-thread-1] threadpool.DefaultScheduler:117 : CubingJob{id=c6709f0b-8858-4e66-a4c2-320ebc70a2e3, name=kylin_sales_cube - 20120101000000_20140201000000 - MERGE - GMT+08:00 2017-05-25 16:51:30, state=READY} scheduled
2017-05-25 17:28:07,075 INFO [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.AbstractExecutable:110 : Executing AbstractExecutable (kylin_sales_cube - 20120101000000_20140201000000 - MERGE - GMT+08:00 2017-05-25 16:51:30)
2017-05-25 17:28:07,078 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] dao.ExecutableDao:217 : updating job output, id: c6709f0b-8858-4e66-a4c2-320ebc70a2e3
2017-05-25 17:28:07,083 INFO [pool-9-thread-1] threadpool.DefaultScheduler:124 : Job Fetcher: 0 should running, 1 actual running, 0 stopped, 1 ready, 19 already succeed, 0 error, 11 discarded, 0 others
2017-05-25 17:28:07,083 INFO [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.ExecutableManager:389 : job id:c6709f0b-8858-4e66-a4c2-320ebc70a2e3 from READY to RUNNING
2017-05-25 17:28:07,105 INFO [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.AbstractExecutable:110 : Executing AbstractExecutable (Garbage Collection on HDFS)
2017-05-25 17:28:07,106 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] dao.ExecutableDao:217 : updating job output, id: c6709f0b-8858-4e66-a4c2-320ebc70a2e3-08
2017-05-25 17:28:07,111 INFO [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.ExecutableManager:389 : job id:c6709f0b-8858-4e66-a4c2-320ebc70a2e3-08 from READY to RUNNING
2017-05-25 17:28:07,154 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] steps.HDFSPathGarbageCollectionStep:78 : Drop HDFS path on FileSystem: hdfs://cdh5
2017-05-25 17:28:07,217 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] steps.HDFSPathGarbageCollectionStep:90 : HDFS path hdfs:///kylin/kylin_metadata/kylin-a11d510f-d8a5-45c1-b430-bc7def851432 not exists.
2017-05-25 17:28:07,249 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] steps.HDFSPathGarbageCollectionStep:90 : HDFS path hdfs:///kylin/kylin_metadata/kylin-0c1ed2d0-f595-4f58-aaea-2dbe7b41a550 not exists.
2017-05-25 17:28:07,320 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] steps.HDFSPathGarbageCollectionStep:78 : Drop HDFS path on FileSystem: hdfs://cdh5-mini
2017-05-25 17:28:07,324 ERROR [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.AbstractExecutable:126 : error running Executable: HDFSPathGarbageCollectionStep{id=c6709f0b-8858-4e66-a4c2-320ebc70a2e3-08, name=Garbage Collection on HDFS, state=RUNNING}
2017-05-25 17:28:07,326 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] dao.ExecutableDao:217 : updating job output, id: c6709f0b-8858-4e66-a4c2-320ebc70a2e3-08
2017-05-25 17:28:07,331 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] dao.ExecutableDao:217 : updating job output, id: c6709f0b-8858-4e66-a4c2-320ebc70a2e3-08
2017-05-25 17:28:07,334 INFO [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.ExecutableManager:389 : job id:c6709f0b-8858-4e66-a4c2-320ebc70a2e3-08 from RUNNING to ERROR
2017-05-25 17:28:07,335 ERROR [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.AbstractExecutable:126 : error running Executable: CubingJob{id=c6709f0b-8858-4e66-a4c2-320ebc70a2e3, name=kylin_sales_cube - 20120101000000_20140201000000 - MERGE - GMT+08:00 2017-05-25 16:51:30, state=RUNNING}
2017-05-25 17:28:07,337 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] dao.ExecutableDao:217 : updating job output, id: c6709f0b-8858-4e66-a4c2-320ebc70a2e3
2017-05-25 17:28:07,342 DEBUG [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] dao.ExecutableDao:217 : updating job output, id: c6709f0b-8858-4e66-a4c2-320ebc70a2e3
2017-05-25 17:28:07,344 INFO [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.ExecutableManager:389 : job id:c6709f0b-8858-4e66-a4c2-320ebc70a2e3 from RUNNING to ERROR
2017-05-25 17:28:07,345 WARN [Job c6709f0b-8858-4e66-a4c2-320ebc70a2e3-128] execution.AbstractExecutable:258 : no need to send email, user list is empty
2017-05-25 17:28:07,346 ERROR [pool-10-thread-1] threadpool.DefaultScheduler:146 : ExecuteException job:c6709f0b-8858-4e66-a4c2-320ebc70a2e3
org.apache.kylin.job.exception.ExecuteException: org.apache.kylin.job.exception.ExecuteException: java.lang.IllegalArgumentException: Wrong FS: hdfs:/kylin/kylin_metadata/kylin-a11d510f-d8a5-45c1-b430-bc7def851432, expected: hdfs://cdh5-mini
at org.apache.kylin.job.execution.AbstractExecutable.execute(AbstractExecutable.java:134)
at org.apache.kylin.job.impl.threadpool.DefaultScheduler$JobRunner.run(DefaultScheduler.java:142)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:745)
Caused by: org.apache.kylin.job.exception.ExecuteException: java.lang.IllegalArgumentException: Wrong FS: hdfs:/kylin/kylin_metadata/kylin-a11d510f-d8a5-45c1-b430-bc7def851432, expected: hdfs://cdh5-mini
at org.apache.kylin.job.execution.AbstractExecutable.execute(AbstractExecutable.java:134)
at org.apache.kylin.job.execution.DefaultChainedExecutable.doWork(DefaultChainedExecutable.java:64)
at org.apache.kylin.job.execution.AbstractExecutable.execute(AbstractExecutable.java:124)
... 4 more
Caused by: java.lang.IllegalArgumentException: Wrong FS: hdfs:/kylin/kylin_metadata/kylin-a11d510f-d8a5-45c1-b430-bc7def851432, expected: hdfs://cdh5-mini
at org.apache.hadoop.fs.FileSystem.checkPath(FileSystem.java:658)
at org.apache.hadoop.hdfs.DistributedFileSystem.getPathName(DistributedFileSystem.java:194)
at org.apache.hadoop.hdfs.DistributedFileSystem.access$000(DistributedFileSystem.java:106)
at org.apache.hadoop.hdfs.DistributedFileSystem$19.doCall(DistributedFileSystem.java:1215)
at org.apache.hadoop.hdfs.DistributedFileSystem$19.doCall(DistributedFileSystem.java:1211)
at org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81)
at org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1211)
at org.apache.hadoop.fs.FileSystem.exists(FileSystem.java:1413)
at org.apache.kylin.storage.hbase.steps.HDFSPathGarbageCollectionStep.dropHdfsPathOnCluster(HDFSPathGarbageCollectionStep.java:85)
at org.apache.kylin.storage.hbase.steps.HDFSPathGarbageCollectionStep.doWork(HDFSPathGarbageCollectionStep.java:65)
at org.apache.kylin.job.execution.AbstractExecutable.execute(AbstractExecutable.java:124)
... 6 more
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
Re: [jira] [Created] (KYLIN-2648) Encounter cube merge error when
deploy kylin on stand alone hbase cluster
Posted by Li Yang <li...@apache.org>.
Let's avoid due post and keep discussion in JIRA. Thanks.
Yang
On Mon, May 29, 2017 at 2:52 PM, suheng.cloud <pe...@hotmail.com>
wrote:
> Hi, shaofeng & yang
> I try to modiry property "kylin.env.hdfs-working-dir" to full qualified
> path
> " hdfs://cdh5/kylin-83-hadoop",
> and found it can't work around.
> The error stacktrace is similar as before:
> "Wrong FS:
> hdfs://cdh5/kylin-83-hadoop/kylin_metadata/kylin-9e6272e3-
> d813-481d-82c0-182b85a0ad8f,
> expected: hdfs://cdh5-mini"
>
> I checked kylin source, in HDFSPathGarbageCollectionStep.java func doWork:
>
> *dropHdfsPathOnCluster(toDeletePaths, HadoopUtil.getWorkingFileSystem());
> if
> (StringUtils.isNotEmpty(context.getConfig().getHBaseClusterFs())) {
> dropHdfsPathOnCluster(toDeletePaths,
> FileSystem.get(HBaseConnection.getCurrentHBaseConfiguration()));
> }*
> As long as the property kylin.storage.hbase.cluster-fs(which I set to
> hdfs://cdh5-mini) has been set,
> Kylin will always try to find and delete the "toDeletePaths" on hbase
> cluster,and if we use full qualified path like
> "hdfs://cdh5/kylin-83-hadoop",in function dropHdfsPathOnCluster, exception
> will be thrown out by the checkpath function since the authority of
> hdfs://cdh5/kylin-83-hadoop/xxx was not equal to the authority of
> hdfs://cdh5-mini".So I think it should be a relative path,so kylin check on
> both hadoop cluster?
>
> Do we have some other work around way? Or what your standard deploy method
> when hbase cluster seperate from main cluster?
>
> I was trying to check if kylin can meet our olap demand,and stuck on this
> for a long time, really need your help, thank you!
>
> full log as follows:
> err_stacktrace.txt
> <http://apache-kylin.74782.x6.nabble.com/file/n8121/err_stacktrace.txt>
>
> Best wishes!
>
> --
> View this message in context: http://apache-kylin.74782.x6.
> nabble.com/jira-Created-KYLIN-2648-Encounter-cube-merge-
> error-when-deploy-kylin-on-stand-alone-hbase-cluster-tp8096p8121.html
> Sent from the Apache Kylin mailing list archive at Nabble.com.
>
Re: [jira] [Created] (KYLIN-2648) Encounter cube merge error when
deploy kylin on stand alone hbase cluster
Posted by "suheng.cloud" <pe...@hotmail.com>.
Hi, shaofeng & yang
I try to modiry property "kylin.env.hdfs-working-dir" to full qualified path
" hdfs://cdh5/kylin-83-hadoop",
and found it can't work around.
The error stacktrace is similar as before:
"Wrong FS:
hdfs://cdh5/kylin-83-hadoop/kylin_metadata/kylin-9e6272e3-d813-481d-82c0-182b85a0ad8f,
expected: hdfs://cdh5-mini"
I checked kylin source, in HDFSPathGarbageCollectionStep.java func doWork:
*dropHdfsPathOnCluster(toDeletePaths, HadoopUtil.getWorkingFileSystem());
if
(StringUtils.isNotEmpty(context.getConfig().getHBaseClusterFs())) {
dropHdfsPathOnCluster(toDeletePaths,
FileSystem.get(HBaseConnection.getCurrentHBaseConfiguration()));
}*
As long as the property kylin.storage.hbase.cluster-fs(which I set to
hdfs://cdh5-mini) has been set,
Kylin will always try to find and delete the "toDeletePaths" on hbase
cluster,and if we use full qualified path like
"hdfs://cdh5/kylin-83-hadoop",in function dropHdfsPathOnCluster, exception
will be thrown out by the checkpath function since the authority of
hdfs://cdh5/kylin-83-hadoop/xxx was not equal to the authority of
hdfs://cdh5-mini".So I think it should be a relative path,so kylin check on
both hadoop cluster?
Do we have some other work around way? Or what your standard deploy method
when hbase cluster seperate from main cluster?
I was trying to check if kylin can meet our olap demand,and stuck on this
for a long time, really need your help, thank you!
full log as follows:
err_stacktrace.txt
<http://apache-kylin.74782.x6.nabble.com/file/n8121/err_stacktrace.txt>
Best wishes!
--
View this message in context: http://apache-kylin.74782.x6.nabble.com/jira-Created-KYLIN-2648-Encounter-cube-merge-error-when-deploy-kylin-on-stand-alone-hbase-cluster-tp8096p8121.html
Sent from the Apache Kylin mailing list archive at Nabble.com.