Progress: I found the very latest r9 patch on the Apache Incubation board.
 Unlike most of the other r1-r8 patches, r9 applies cleanly to trunk.  With
some work, I was able to build that and do "make hadoop-2.0.0-mr1-cdh4.1.2"
in the "hadoop" subdirectory.  The build process downloaded the needed
tarball, built Hadoop 2 with the Mesos patch, and then successfully ran
some Hadoop self tests.

Now I will install this Mesos and Hadoop 2 on a test machine and see if I
can get it all to work in the Mesos "standalone" configuration.

Craig

On Fri, Mar 1, 2013 at 10:22 AM, Craig Vanderborgh <
[email protected]> wrote:

> Thanks for the reply, Vinod..
>
> We are running 0.9.0-incubating downloaded from here:
>
>
> http://www.apache.org/dyn/closer.cgi/incubator/mesos/mesos-0.9.0-incubating/
>
> The Hadoop I'm trying to run is v0.20.205 that's included with this Mesos
> distribution.  But we WANT to run hadoop-2.0.0-mr1-cdh4.1.2.  I've got a
> patch for this, but not sure exactly what I should use as the starting
> point.  Can you give me a few pointers on how to proceed?  We want the
> hadoop-2.0.0-mr-cdh4.1.2 version, and we'll build/run anything required to
> get it.
>
> Craig
>
>
> On Fri, Mar 1, 2013 at 9:16 AM, Vinod Kone <[email protected]> wrote:
>
>> Hey Craig,
>>
>> Which version of Mesos are you running? We recently (~1 month) revamped
>> our Hadoop port and unfortunately didn't update the wiki.
>>
>> @vinodkone
>> Sent from my mobile
>>
>> On Mar 1, 2013, at 7:48 AM, Craig Vanderborgh <[email protected]>
>> wrote:
>>
>> > I'm getting close.  I've got the namenode formatted and HDFS running.  I
>> > started hdfs with $HADOOP_HOME/bin/start-dfs.sh.  Basic HDFS operations
>> > seem to be working.  Now it's time to start the jobtracker - I have been
>> > following the instructions here:
>> >
>> > https://github.com/mesos/mesos/wiki/Running-Hadoop-on-Mesos
>> >
>> > The instructions say "Launch a JobTracker with bin/hadoop jobtracker (do
>> > not use bin/start-mapred.sh)...
>> >
>> > So here's what happens:
>> >
>> > [craigv@sleuth-mesos-01 bin]$ hadoop jobtracker
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: STARTUP_MSG:
>> > /************************************************************
>> > STARTUP_MSG: Starting JobTracker
>> > STARTUP_MSG:   host = sleuth-mesos-01.lab.webapps.rr.com/71.74.48.186
>> > STARTUP_MSG:   args = []
>> > STARTUP_MSG:   version = 0.20.205.1
>> > STARTUP_MSG:   build =  -r ; compiled by 'craigv' on Thu Feb 28 22:59:28
>> > UTC 2013
>> > ************************************************************/
>> > 13/03/01 00:23:02 INFO impl.MetricsConfig: loaded properties from
>> > hadoop-metrics2.properties
>> > 13/03/01 00:23:02 INFO impl.MetricsSourceAdapter: MBean for source
>> > MetricsSystem,sub=Stats registered.
>> > 13/03/01 00:23:02 INFO impl.MetricsSystemImpl: Scheduled snapshot
>> period at
>> > 10 second(s).
>> > 13/03/01 00:23:02 INFO impl.MetricsSystemImpl: JobTracker metrics system
>> > started
>> > 13/03/01 00:23:02 INFO impl.MetricsSourceAdapter: MBean for source
>> > QueueMetrics,q=default registered.
>> > 13/03/01 00:23:02 INFO impl.MetricsSourceAdapter: MBean for source ugi
>> > registered.
>> > 13/03/01 00:23:02 INFO delegation.AbstractDelegationTokenSecretManager:
>> > Updating the current master key for generating delegation tokens
>> > 13/03/01 00:23:02 INFO delegation.AbstractDelegationTokenSecretManager:
>> > Starting expired delegation token remover thread,
>> > tokenRemoverScanInterval=60 min(s)
>> > 13/03/01 00:23:02 INFO delegation.AbstractDelegationTokenSecretManager:
>> > Updating the current master key for generating delegation tokens
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: Scheduler configured with
>> > (memSizeForMapSlotOnJT, memSizeForReduceSlotOnJT,
>> limitMaxMemForMapTasks,
>> > limitMaxMemForReduceTasks) (-1, -1, -1, -1)
>> > 13/03/01 00:23:02 INFO util.HostsFileReader: Refreshing hosts
>> > (include/exclude) list
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: Starting jobtracker with
>> owner as
>> > craigv
>> > 13/03/01 00:23:02 INFO ipc.Server: Starting SocketReader
>> > 13/03/01 00:23:02 INFO impl.MetricsSourceAdapter: MBean for source
>> > RpcDetailedActivityForPort54311 registered.
>> > 13/03/01 00:23:02 INFO impl.MetricsSourceAdapter: MBean for source
>> > RpcActivityForPort54311 registered.
>> > 13/03/01 00:23:02 INFO mortbay.log: Logging to
>> > org.slf4j.impl.Log4jLoggerAdapter(org.mortbay.log) via
>> > org.mortbay.log.Slf4jLog
>> > 13/03/01 00:23:02 INFO http.HttpServer: Added global filtersafety
>> > (class=org.apache.hadoop.http.HttpServer$QuotingInputFilter)
>> > 13/03/01 00:23:02 INFO http.HttpServer: Port returned by
>> > webServer.getConnectors()[0].getLocalPort() before open() is -1. Opening
>> > the listener on 50030
>> > 13/03/01 00:23:02 INFO http.HttpServer: listener.getLocalPort() returned
>> > 50030 webServer.getConnectors()[0].getLocalPort() returned 50030
>> > 13/03/01 00:23:02 INFO http.HttpServer: Jetty bound to port 50030
>> > 13/03/01 00:23:02 INFO mortbay.log: jetty-6.1.26
>> > 13/03/01 00:23:02 INFO mortbay.log: Started
>> > [email protected]:50030
>> > 13/03/01 00:23:02 INFO impl.MetricsSourceAdapter: MBean for source jvm
>> > registered.
>> > 13/03/01 00:23:02 INFO impl.MetricsSourceAdapter: MBean for source
>> > JobTrackerMetrics registered.
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: JobTracker up at: 54311
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: JobTracker webserver: 50030
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: Cleaning up the system
>> directory
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: History server being
>> initialized
>> > in embedded mode
>> > 13/03/01 00:23:02 INFO mapred.JobHistoryServer: Started job history
>> server
>> > at: localhost:50030
>> > 13/03/01 00:23:02 INFO mapred.JobTracker: Job History Server web
>> address:
>> > localhost:50030
>> > 13/03/01 00:23:02 INFO mapred.CompletedJobStatusStore: Completed job
>> store
>> > is inactive
>> > 13/03/01 00:23:03 INFO mapred.MesosScheduler: Starting MesosScheduler
>> > java: ./common/try.hpp:77: T Try<T>::get() const [with T =
>> > mesos::internal::MasterDetector*]: Assertion `state == SOME' failed.
>> > Aborted
>> >
>> > Close but no cigar.  The Mesos code in this proximity:
>> >
>> > src/common/try.hpp
>> >
>> > Try<T>& operator = (const Try<T>& that)
>> >  {
>> >    if (this != &that) {
>> >      state = that.state;
>> >      if (that.t != NULL) {
>> >        t = new T(*that.t);
>> >      } else {
>> >        t = NULL;
>> >      }
>> >      message = that.message;
>> >    }
>> >
>> >    return *this;
>> >  }
>> >
>> >  bool isSome() const { return state == SOME; }
>> >  bool isError() const { return state == ERROR; }
>> >
>> >  T get() const { assert(state == SOME); return *t; }
>> >
>> >  std::string error() const { assert(state == ERROR); return message; }
>> >
>> > What's wrong here?  Obviously it can't start MesosScheduler, but why?
>> >
>> > Here is my mapred-site.xml:
>> >
>> > <?xml version="1.0"?>
>> > <?xml-stylesheet type="text/xsl" href="configuration.xsl"?>
>> >
>> > <!-- Put site-specific property overrides in this file. -->
>> >
>> > <configuration>
>> >  <property>
>> >    <name>mapred.job.tracker</name>
>> >    <value>localhost:54311</value>
>> >  </property> -->
>> >  <property>
>> >    <name>mapred.jobtracker.taskScheduler</name>
>> >    <value>org.apache.hadoop.mapred.MesosScheduler</value>
>> >  </property>
>> >  <property>
>> >    <name>mapred.mesos.master</name>
>> >    <value>mesos://sleuth-mesos-01:5050</value>
>> >  </property>
>> > </configuration>
>> >
>> > THANKS!
>> > Craig
>>
>
>

Reply via email to