As a follow-up, registration failure seems to have been based on using the wrong mesosMaster IP address or format. I changed it to "zk://10.0.2.15:2181/mesos” following syntax from this list archive: https://mail-archives.apache.org/mod_mbox/myriad-dev/201602.mbox/%3c1519159574.1366234.1456242921704.javamail.ya...@mail.yahoo.com%3e
The mesos master now shows MyriadAlpha as an active framework. Still, the myriad tasks list shows the default medium as a pending task, so there still seems to be a problem. The Mesos slave does not show any frameworks or completed frameworks. Again, just trying to test-drive the vanilla vagrant install. Happy to provide notes of what works and doesn’t if anyone wants to update the vagrant install docs to Myriad 0.2.0: https://cwiki.apache.org/confluence/display/MYRIAD/Installing+using+Vagrant Further detail on users: step 1 seems best to complete as the vagrant user remaining steps seem to need to be completed as the hadoop user (i.e. hduser) Regards, David > On Jul 26, 2016, at 9:27 AM, Reno, David <david_r...@comcast.com> wrote: > > Hi Myriaders, > > Sorry if I’m reaching out to the wrong alias or help, this is all I see. I’m > getting stuck with the myriad install with vagrant. The wiki seem to assume > 0.1.0 though I’ve cloned the latest 0.2.0 release from github. > > I’m following these instructions: > https://cwiki.apache.org/confluence/display/MYRIAD/Installing+using+Vagrant > > Step 1 seems to go fine and I can open the HDFS name node and mesos master > http ports and see the pages showing active/started. Step 2 starts go to a > little sideways as it references “myriad-executor-0.1.0.jar” which seems to > be replaced by “myriad-executor-0.2.0.jar” which I use instead. Step 3 asks > for minimum configuration changes which seem to already be completed. > However, I change the line: > path: > file://localhost/usr/local/libexec/mesos/myriad-executor-runnable-0.1.0.jar > to: > path: > file:///usr/local/hadoop/share/hadoop/yarn/lib/myriad-executor-0.2.0.jar > > For step 4, I add all properties listed to the yarn-site.xml file. I then > launch the resource manager using the “yarn-daemon.sh start resourcemanager” > command. > > At this point, I can load the http://10.141.141.20:8192 port and see the > myriad about and API page but the http://10.141.141.20:5050/#/frameworks page > does not show myriad or hadoop as an active framework. I use the myriad flex > tab to “flex up” a small server, it appears as a pending task, but stays > pending and mesos frameworks don’t change. > > Interesting lines from > /usr/local/hadoop/logs/yarn-hduser-resourcemanager-vagrant-ubuntu-trusty-64.out > include the following: > I0726 13:01:41.358747 15817 sched.cpp:164] Version: 0.24.1 > I0726 13:01:41.361140 15847 sched.cpp:262] New master detected at > master@10.0.2.15:5050 > I0726 13:01:41.361538 15847 sched.cpp:272] No credentials provided. > Attempting to register without authentication > E0726 13:01:41.362741 15852 socket.hpp:174] Shutdown failed on fd=231: > Transport endpoint is not connected [107] > E0726 13:01:41.363302 15852 socket.hpp:174] Shutdown failed on fd=231: > Transport endpoint is not connected [107] > E0726 13:01:41.396867 15852 socket.hpp:174] Shutdown failed on fd=231: > Transport endpoint is not connected [107] > Jul 26, 2016 1:01:41 PM com.google.inject.servlet.GuiceFilter setPipeline > WARNING: Multiple Servlet injectors detected. This is a warning indicating > that you have more than one GuiceFilter running in your web application. If > this is deliberate, you may safely ignore this message. If this is NOT > deliberate however, your application may not work as expected. > E0726 13:01:44.780588 15852 socket.hpp:174] Shutdown failed on fd=275: > Transport endpoint is not connected [107] > E0726 13:01:51.604310 15852 socket.hpp:174] Shutdown failed on fd=275: > Transport endpoint is not connected [107] > E0726 13:02:01.226771 15852 socket.hpp:174] Shutdown failed on fd=275: > Transport endpoint is not connected [107] > E0726 13:02:11.525804 15852 socket.hpp:174] Shutdown failed on fd=277: > Transport endpoint is not connected [107] > Jul 26, 2016 1:02:15 PM > com.sun.jersey.server.wadl.generators.WadlGeneratorJAXBGrammarGenerator$8 > resolve > SEVERE: null > java.lang.IllegalAccessException: Class > com.sun.jersey.server.wadl.generators.WadlGeneratorJAXBGrammarGenerator$8 can > not access a member of class javax.ws.rs.core.Response with modifiers > "protected" > > Any help or suggestions are much appreciated, > David Reno > Systems Architect, Comcast
