You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@livy.apache.org by "Hari Sekhon (JIRA)" <ji...@apache.org> on 2018/12/11 10:04:01 UTC
[jira] [Updated] (LIVY-541) Multiple Livy servers submitting to
Yarn results in LivyException: Session is finished ... No YARN application
is found with tag livy-session-197-uveqmqyj in 300 seconds. Please check
your cluster status, it is may be very busy
[ https://issues.apache.org/jira/browse/LIVY-541?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Hari Sekhon updated LIVY-541:
-----------------------------
Description:
It appears Livy doesn't differentiate sessions properly in Yarn causing errors when running multiple Livy servers behind a load balancer for HA and performance scaling on the same Hadoop cluster.
Each livy server uses monotonically incrementing session IDs with a random suffix but it appears that the random suffix isn't passed to Yarn which results in the following errors on the Livy server which is further behind in session numbers because it appears to find the session with the same number has already finished (submitted earlier by a different user on another Livy server as seen in Yarn RM UI):
{code:java}
org.apache.zeppelin.livy.LivyException: Session 197 is finished, appId: null, log: [ at org.apache.hadoop.fs.FileSystem$Cache.closeAll(FileSystem.java:2887), at org.apache.hadoop.fs.FileSystem$Cache$ClientFinalizer.run(FileSystem.java:2904), at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511), at java.util.concurrent.FutureTask.run(FutureTask.java:266), at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142), at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617), at java.lang.Thread.run(Thread.java:748),
YARN Diagnostics: , java.lang.Exception: No YARN application is found with tag livy-session-197-uveqmqyj in 300 seconds. Please check your cluster status, it is may be very busy., org.apache.livy.utils.SparkYarnApp.org$apache$livy$utils$SparkYarnApp$$getAppIdFromTag(SparkYarnApp.scala:182) org.apache.livy.utils.SparkYarnApp$$anonfun$1$$anonfun$4.apply(SparkYarnApp.scala:239) org.apache.livy.utils.SparkYarnApp$$anonfun$1$$anonfun$4.apply(SparkYarnApp.scala:236) scala.Option.getOrElse(Option.scala:120) org.apache.livy.utils.SparkYarnApp$$anonfun$1.apply$mcV$sp(SparkYarnApp.scala:236) org.apache.livy.Utils$$anon$1.run(Utils.scala:94)]
at org.apache.zeppelin.livy.BaseLivyInterpreter.createSession(BaseLivyInterpreter.java:300)
at org.apache.zeppelin.livy.BaseLivyInterpreter.initLivySession(BaseLivyInterpreter.java:184)
at org.apache.zeppelin.livy.LivySharedInterpreter.open(LivySharedInterpreter.java:57)
at org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:69)
at org.apache.zeppelin.livy.BaseLivyInterpreter.getLivySharedInterpreter(BaseLivyInterpreter.java:165)
at org.apache.zeppelin.livy.BaseLivyInterpreter.open(BaseLivyInterpreter.java:139)
at org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:69)
at org.apache.zeppelin.interpreter.remote.RemoteInterpreterServer$InterpretJob.jobRun(RemoteInterpreterServer.java:493)
at org.apache.zeppelin.scheduler.Job.run(Job.java:175)
at org.apache.zeppelin.scheduler.FIFOScheduler$1.run(FIFOScheduler.java:139)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
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:748)
{code}
was:
It appears Livy doesn't differentiate sessions properly in Yarn causing errors when running multiple Livy servers behind a load balancer for HA and performance scaling on the same Hadoop cluster.
Each livy server uses monotonically incrementing session IDs with a hash suffix but it appears that the hash suffix isn't passed to Yarn which results in the following errors on the Livy server which is further behind in session numbers because it appears to find the session with the same number has already finished (submitted earlier by a different user on another Livy server as seen in Yarn RM UI):
{code:java}
org.apache.zeppelin.livy.LivyException: Session 197 is finished, appId: null, log: [ at org.apache.hadoop.fs.FileSystem$Cache.closeAll(FileSystem.java:2887), at org.apache.hadoop.fs.FileSystem$Cache$ClientFinalizer.run(FileSystem.java:2904), at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511), at java.util.concurrent.FutureTask.run(FutureTask.java:266), 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:748),
YARN Diagnostics: , java.lang.Exception: No YARN application is found with tag livy-session-197-uveqmqyj in 300 seconds. Please check your cluster status, it is may be very busy., org.apache.livy.utils.SparkYarnApp.org$apache$livy$utils$SparkYarnApp$$getAppIdFromTag(SparkYarnApp.scala:182) org.apache.livy.utils.SparkYarnApp$$anonfun$1$$anonfun$4.apply(SparkYarnApp.scala:239) org.apache.livy.utils.SparkYarnApp$$anonfun$1$$anonfun$4.apply(SparkYarnApp.scala:236) scala.Option.getOrElse(Option.scala:120) org.apache.livy.utils.SparkYarnApp$$anonfun$1.apply$mcV$sp(SparkYarnApp.scala:236) org.apache.livy.Utils$$anon$1.run(Utils.scala:94)]
at org.apache.zeppelin.livy.BaseLivyInterpreter.createSession(BaseLivyInterpreter.java:300)
at org.apache.zeppelin.livy.BaseLivyInterpreter.initLivySession(BaseLivyInterpreter.java:184)
at org.apache.zeppelin.livy.LivySharedInterpreter.open(LivySharedInterpreter.java:57)
at org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:69)
at org.apache.zeppelin.livy.BaseLivyInterpreter.getLivySharedInterpreter(BaseLivyInterpreter.java:165)
at org.apache.zeppelin.livy.BaseLivyInterpreter.open(BaseLivyInterpreter.java:139)
at org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:69)
at org.apache.zeppelin.interpreter.remote.RemoteInterpreterServer$InterpretJob.jobRun(RemoteInterpreterServer.java:493)
at org.apache.zeppelin.scheduler.Job.run(Job.java:175)
at org.apache.zeppelin.scheduler.FIFOScheduler$1.run(FIFOScheduler.java:139)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
at java.lang.Thread.run(Thread.java:748)
{code}
> Multiple Livy servers submitting to Yarn results in LivyException: Session is finished ... No YARN application is found with tag livy-session-197-uveqmqyj in 300 seconds. Please check your cluster status, it is may be very busy
> -----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: LIVY-541
> URL: https://issues.apache.org/jira/browse/LIVY-541
> Project: Livy
> Issue Type: Bug
> Components: Server
> Affects Versions: 0.5.0
> Environment: Hortonworks HDP 2.6
> Reporter: Hari Sekhon
> Priority: Critical
>
> It appears Livy doesn't differentiate sessions properly in Yarn causing errors when running multiple Livy servers behind a load balancer for HA and performance scaling on the same Hadoop cluster.
> Each livy server uses monotonically incrementing session IDs with a random suffix but it appears that the random suffix isn't passed to Yarn which results in the following errors on the Livy server which is further behind in session numbers because it appears to find the session with the same number has already finished (submitted earlier by a different user on another Livy server as seen in Yarn RM UI):
> {code:java}
> org.apache.zeppelin.livy.LivyException: Session 197 is finished, appId: null, log: [ at org.apache.hadoop.fs.FileSystem$Cache.closeAll(FileSystem.java:2887), at org.apache.hadoop.fs.FileSystem$Cache$ClientFinalizer.run(FileSystem.java:2904), at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511), at java.util.concurrent.FutureTask.run(FutureTask.java:266), at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142), at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617), at java.lang.Thread.run(Thread.java:748),
> YARN Diagnostics: , java.lang.Exception: No YARN application is found with tag livy-session-197-uveqmqyj in 300 seconds. Please check your cluster status, it is may be very busy., org.apache.livy.utils.SparkYarnApp.org$apache$livy$utils$SparkYarnApp$$getAppIdFromTag(SparkYarnApp.scala:182) org.apache.livy.utils.SparkYarnApp$$anonfun$1$$anonfun$4.apply(SparkYarnApp.scala:239) org.apache.livy.utils.SparkYarnApp$$anonfun$1$$anonfun$4.apply(SparkYarnApp.scala:236) scala.Option.getOrElse(Option.scala:120) org.apache.livy.utils.SparkYarnApp$$anonfun$1.apply$mcV$sp(SparkYarnApp.scala:236) org.apache.livy.Utils$$anon$1.run(Utils.scala:94)]
> at org.apache.zeppelin.livy.BaseLivyInterpreter.createSession(BaseLivyInterpreter.java:300)
> at org.apache.zeppelin.livy.BaseLivyInterpreter.initLivySession(BaseLivyInterpreter.java:184)
> at org.apache.zeppelin.livy.LivySharedInterpreter.open(LivySharedInterpreter.java:57)
> at org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:69)
> at org.apache.zeppelin.livy.BaseLivyInterpreter.getLivySharedInterpreter(BaseLivyInterpreter.java:165)
> at org.apache.zeppelin.livy.BaseLivyInterpreter.open(BaseLivyInterpreter.java:139)
> at org.apache.zeppelin.interpreter.LazyOpenInterpreter.open(LazyOpenInterpreter.java:69)
> at org.apache.zeppelin.interpreter.remote.RemoteInterpreterServer$InterpretJob.jobRun(RemoteInterpreterServer.java:493)
> at org.apache.zeppelin.scheduler.Job.run(Job.java:175)
> at org.apache.zeppelin.scheduler.FIFOScheduler$1.run(FIFOScheduler.java:139)
> at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
> at java.util.concurrent.FutureTask.run(FutureTask.java:266)
> at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
> at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
> 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:748)
> {code}
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)