[
https://issues.apache.org/jira/browse/SPARK-13776?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Owen resolved SPARK-13776.
-------------------------------
Resolution: Fixed
Fix Version/s: 2.0.0
Issue resolved by pull request 11615
[https://github.com/apache/spark/pull/11615]
> Web UI is not available after ./sbin/start-master.sh
> ----------------------------------------------------
>
> Key: SPARK-13776
> URL: https://issues.apache.org/jira/browse/SPARK-13776
> Project: Spark
> Issue Type: Bug
> Components: Web UI
> Affects Versions: 1.6.0
> Environment: Solaris 11.3, Oracle SPARC T-5 8 with 1024 hardware
> threads
> Reporter: Erik O'Shaughnessy
> Priority: Minor
> Fix For: 2.0.0
>
>
> The Apache Spark Web UI fails to become available after starting a Spark
> master in stand-alone mode:
> $ ./sbin/start-master.sh
> The log file contains the following:
> {quote}
> cat spark-hadoop-org.apache.spark.deploy.master.Master-1-t5-8-002.out
> Spark Command: /usr/java/bin/java -cp
> /usr/local/spark-1.6.0_nohadoop/conf/:/usr/local/spark-1.6.0_nohadoop/assembly/target/scala-2.10/spark-assembly-1.6.0-hadoop2.2.0.jar:/usr/local/spark-1.6.0_nohadoop/lib_managed/jars/datanucleus-api-jdo-3.2.6.jar:/usr/local/spark-1.6.0_nohadoop/lib_managed/jars/datanucleus-rdbms-3.2.9.jar:/usr/local/spark-1.6.0_nohadoop/lib_managed/jars/datanucleus-core-3.2.10.jar
> -Xms1g -Xmx1g org.apache.spark.deploy.master.Master --ip t5-8-002 --port
> 7077 --webui-port 8080
> ========================================
> 16/01/27 12:00:42 WARN AbstractConnector: insufficient threads configured for
> SelectChannelConnector@0.0.0.0:8080
> 16/01/27 12:00:42 WARN AbstractConnector: insufficient threads configured for
> SelectChannelConnector@t5-8-002:6066
> {quote}
> I did some poking around and it seems that message is coming from Jetty and
> indicates a mismatch between Jetty's default maxThreads configuration and the
> actual number of CPUs available on the hardware (1024). I was not able to
> find a way to successfully change Jetty's configuration at run-time.
> Our work around was to disable CPUs until the WARN messages did not occur in
> the log file, which was when NCPUs = 504.
> I don't know for certain that this is isn't a known problem in Jetty from
> looking at their bug reports, but I wasn't able to locate a Jetty issue that
> described this problem.
> While not specifically an Apache Spark problem, I thought documenting it
> would at least be helpful.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
