[StarCluster] Help needed with running mpich2 plugin using starcluster

[Subbarao Kota]

Hi Justin, Thanks so much for taking time on this.

Everytimg I am starting the cluster using StarCluster I am receiving the errors as in the attached file. I have also attached my config file. Some values are replaced such as access_keys.

1. Why are these errors being received? See attached file (cluster_startup_errors) for errors. Also attached is the crash_report.
2. I included a [plugin mpich2] section in the config file but when starting the cluster with this setting doesn't seem start the plugins. Why is that?

My goal is to run some benchmarks (specifically STREAM (MPI version in C), IOR and NAS MPI) on Amazon EC2 instances. Once I am successful in running these on just 1 instance, my plan is to run them on 2 nodes, 4 nodes, 6 nodes and 8 nodes to evaluate the performance of these benchmarks. And I need to run them with MPICH2.

So far I have been able to run STREAM and IOR benchmarks on t1.micro instance. Now I want to create a small cluster of 2 nodes and run them. I think StarCluster  can help me accomplish it.

Thanks,
Subbarao Kota.


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: cluster_startup_errors.txt
Url: http://mailman.mit.edu/pipermail/starcluster/attachments/20120226/176fcb0f/attachment-0003.txt
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: starcluster_configfile.txt
Url: http://mailman.mit.edu/pipermail/starcluster/attachments/20120226/176fcb0f/attachment-0004.txt
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: crash-report-9647.txt
Url: http://mailman.mit.edu/pipermail/starcluster/attachments/20120226/176fcb0f/attachment-0005.txt
-------------- next part --------------

On Feb 22, 2012, at 11:51 AM, Justin Riley wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Hi Subbarao,
> 
> Which AMI are you using? This error usually occurs because you're
> using an AMI that does not allow root logins.
> 
> In general you should use either the StarCluster supported AMIs or an
> AMI based on the StarCluster supported AMIs. Have a look at the
> following doc on how to customize the StarCluster supported AMIs:
> 
> http://web.mit.edu/star/cluster/docs/latest/manual/create_new_ami.html
> 
> Also, do you still have the 'crash report' that was generated when you
> got this error? It would be useful to take a look at this as well if
> it's still around....
> 
> ~Justin
> 
> On 02/20/2012 02:57 PM, Subbarao Kota wrote:
>> Hello, I received the below error when attempted to run the mpich2
>> plugin on a 2-node cluster of t1.micro instance via starcluster.
>> Can you please let me know what needs to be fixed or corrected?
>> what additional information you will need to help.
>> 
>> Please advise. Thanks.
>> 
>> ====
>> 
>> StarCluster - (http://web.mit.edu/starcluster) (v. 0.93.1) Software
>> Tools for Academics and Researchers (STAR) Please submit bug
>> reports to starcluster at mit.edu
>> 
>>>>> Running plugin mpich2 Creating MPICH2 hosts file
>> !!! ERROR - Error occurred while running plugin 'mpich2': Traceback
>> (most recent call last): File
>> "/Library/Python/2.7/site-packages/StarCluster-0.93.1-py2.7.egg/starcluster/cluster.py",
>> line 1482, in run_plugin func(*args) File
>> "/Library/Python/2.7/site-packages/StarCluster-0.93.1-py2.7.egg/starcluster/plugins/mpich2.py",
>> line 33, in run mpich2_hosts =
>> master.ssh.remote_file(self.MPICH2_HOSTS, 'w') File
>> "/Library/Python/2.7/site-packages/StarCluster-0.93.1-py2.7.egg/starcluster/ssh.py",
>> line 290, in remote_file rfile = self.sftp.open(file, mode) File
>> "/Library/Python/2.7/site-packages/StarCluster-0.93.1-py2.7.egg/starcluster/ssh.py",
>> line 180, in sftp self._sftp =
>> paramiko.SFTPClient.from_transport(self.transport) File
>> "build/bdist.macosx-10.7-intel/egg/paramiko/sftp_client.py", line
>> 106, in from_transport return cls(chan) File
>> "build/bdist.macosx-10.7-intel/egg/paramiko/sftp_client.py", line
>> 87, in __init__ server_version = self._send_version() File
>> "build/bdist.macosx-10.7-intel/egg/paramiko/sftp.py", line 108, in
>> _send_version t, data = self._read_packet() File
>> "build/bdist.macosx-10.7-intel/egg/paramiko/sftp.py", line 179, in
>> _read_packet raise SFTPError('Garbage packet received') SFTPError:
>> Garbage packet received 
>> _______________________________________________ StarCluster mailing
>> list StarCluster at mit.edu 
>> http://mailman.mit.edu/mailman/listinfo/starcluster
> 
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v2.0.17 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
> 
> iEYEARECAAYFAk9FHPUACgkQ4llAkMfDcrluFwCfcrRe9sW8RRHCmHhK9z4BRp6a
> 5eAAn0nVlMsvg1OQY+RjELzq3Vo5krnQ
> =Eby5
> -----END PGP SIGNATURE-----