Quantcast

[IPython-User] ipcluster in ssh mode in ipython 0.13

classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

[IPython-User] ipcluster in ssh mode in ipython 0.13

M. Wimmer
Dear all,

Having freshly installed ipython 0.13 I tried to get an ipython parallel cluster
running via ssh. For this I followed precisely the documentation and did

 ipython profile create --parallel --profile=ssh

and modified, according to the docs, in ipcluster_config.py the lines

 c.IPClusterEngines.engine_launcher_class = 'SSHEngineSetLauncher'

 c.SSHEngineSetLauncher.engines = {'xxx049': 1}

Trying to start a cluster then misarebly fails

 ipcluster start --profile=ssh --debug

and I've added the debug output below. The failed `scp`-commands I can avoid by
writing in ipcluster_config.py

 c.SSHLauncher.to_send = []
 c.SSHLauncher.to_fetch = []

As you can see from the logfile below, the engines are not started because of
arguments (""--profile ...") being added to the python command. I can change the
engine_cmd to `ipengine` but then I ran into other troubles, so I wanted to
discuss this step by step.

What am I doing wrong? What should be added to the docs?

- Michael

Logfile:
---------------------------------------------------------
2012-07-18 16:40:33,001.001 [IPClusterStart] Config changed:
2012-07-18 16:40:33,001.001 [IPClusterStart] {'Application': {'log_level': 10},
 'BaseIPythonApplication': {'profile': u'ssh'}}
2012-07-18 16:40:33,003.003 [IPClusterStart] Using existing profile dir:
 u'/home/wimmer/.ipython/profile_ssh'
2012-07-18 16:40:33,003.003 [IPClusterStart] Searching path [u'/home/wimmer',
 u'/home/wimmer/.ipython/profile_ssh'] for config files
2012-07-18 16:40:33,003.003 [IPClusterStart] Attempting to load config file:
 ipython_config.py
2012-07-18 16:40:33,005.005 [IPClusterStart] Loaded config file:
 /home/wimmer/.ipython/profile_ssh/ipython_config.py
2012-07-18 16:40:33,005.005 [IPClusterStart] Attempting to load config file:
 ipcluster_config.py
2012-07-18 16:40:33,007.007 [IPClusterStart] Loaded config file:
 /home/wimmer/.ipython/profile_ssh/ipcluster_config.py
2012-07-18 16:40:33,009.009 [IPClusterStart] Config changed:
2012-07-18 16:40:33,009.009 [IPClusterStart] {'Application': {'log_level': 10},
 'BaseIPythonApplication': {'profile': u'ssh'}, 'IPClusterEngines':
 {'engine_launcher_class': 'SSHEngineSetLauncher'}, 'SSHEngineSetLauncher':
 {'engines': {'xxx049': 2}}, 'ProfileDir': {}}
2012-07-18 16:40:33.012 [IPClusterStart] Starting ipcluster with [daemon=False]
2012-07-18 16:40:33.012 [IPClusterStart] Creating pid file:
 /home/wimmer/.ipython/profile_ssh/pid/ipcluster.pid
2012-07-18 16:40:33.013 [IPClusterStart] Starting Controller with
 LocalControllerLauncher
2012-07-18 16:40:33.013 [IPClusterStart] Starting LocalControllerLauncher:
 ['/usr/bin/python', '-c', 'from IPython.parallel.apps.ipcontrollerapp import
 launch_new_instance; launch_new_instance()', '--profile-dir',
 u'/home/wimmer/.ipython/profile_ssh', '--cluster-id', u'', '--log-to-file',
 '--log-level=20']
2012-07-18 16:40:33.020 [IPClusterStart] Process '/usr/bin/python' started:  
 16130
2012-07-18 16:40:33.747 [IPClusterStart] 2012-07-18 16:40:33,746.746
 [IPControllerApp] Using existing profile dir: u'/home/wimmer/.ipython
 /profile_ssh'
2012-07-18 16:40:34.012 [IPClusterStart] Starting 2 Engines with
 SSHEngineSetLauncher
2012-07-18 16:40:34.017 [IPClusterStart] waiting for
 /home/wimmer/.ipython/profile_ssh/security/ipcontroller-client.json
2012-07-18 16:40:35.021 [IPClusterStart] sending
 /home/wimmer/.ipython/profile_ssh/security/ipcontroller-client.json to
 xxx049:security/ipcontroller-client.json
 scp: security/ipcontroller-client.json: No such file or directory
2012-07-18 16:40:36.792 [IPClusterStart] sending
 /home/wimmer/.ipython/profile_ssh/security/ipcontroller-engine.json to
 xxx049:security/ipcontroller-engine.json
 scp: security/ipcontroller-engine.json: No such file or directory
2012-07-18 16:40:37.957 [IPClusterStart] Starting SSHEngineLauncher: ['ssh',
 '-tt', u'xxx049', '/usr/bin/python', '-c', 'from
 IPython.parallel.apps.ipengineapp import launch_new_instance;
 launch_new_instance()', '--profile-dir', u'', '--log-to-file',
 '--log-level=20']
2012-07-18 16:40:37.969 [IPClusterStart] Process 'ssh' started: 16235
2012-07-18 16:40:38.075 [IPClusterStart] Starting SSHEngineLauncher: ['ssh',
 '-tt', u'xxx049', '/usr/bin/python', '-c', 'from
 IPython.parallel.apps.ipengineapp import launch_new_instance;
 launch_new_instance()', '--profile-dir', u'', '--log-to-file',
 '--log-level=20']
2012-07-18 16:40:38.085 [IPClusterStart] Process 'ssh' started: 16237
2012-07-18 16:40:38.086 [IPClusterStart] Process 'engine set' started: [None,  
 None]
2012-07-18 16:40:38.086 [IPClusterStart] 2012-07-18 16:40:34.055 [scheduler]
 Scheduler started [leastload]
2012-07-18 16:40:39.249 [IPClusterStart] bash: -c: line 0: syntax error near
 unexpected token `--profile-dir'
2012-07-18 16:40:39.249 [IPClusterStart] Connection to xxx049 closed.
2012-07-18 16:40:39.250 [IPClusterStart] bash: -c: line 0: `/usr/bin/python -c
 from IPython.parallel.apps.ipengineapp import launch_new_instance;
 launch_new_instance() --profile-dir  --log-to-file --log-level=20'
2012-07-18 16:40:39.273 [IPClusterStart] Process 'ssh' stopped: {'pid': 16235,
 'exit_code': 1}
2012-07-18 16:40:39.451 [IPClusterStart] bash: -c: line 0: syntax error near
 unexpected token `--profile-dir'
2012-07-18 16:40:39.451 [IPClusterStart] Connection to xxx049 closed.
2012-07-18 16:40:39.452 [IPClusterStart] bash: -c: line 0: `/usr/bin/python -c
 from IPython.parallel.apps.ipengineapp import launch_new_instance;
 launch_new_instance() --profile-dir  --log-to-file --log-level=20'
2012-07-18 16:40:39.455 [IPClusterStart] Process 'ssh' stopped: {'pid': 16237,
 'exit_code': 1}
2012-07-18 16:40:39.455 [IPClusterStart] Process 'engine set' stopped:
 {'xxx049/0': {'pid': 16235, 'exit_code': 1}, 'xxx049/1': {'pid': 16237,
 'exit_code': 1}}
2012-07-18 16:40:39.455 [IPClusterStart]
            Engines shutdown early, they probably failed to connect.
           
            Check the engine log files for output.
           
            If your controller and engines are not on the same machine, you probably
            have to instruct the controller to listen on an interface other than
localhost.
           
            You can set this by adding "--ip='*'" to your
ControllerLauncher.controller_args.
           
            Be sure to read our security docs before instructing your controller
to listen on
            a public interface.
           
2012-07-18 16:40:39.456 [IPClusterStart] IPython cluster: stopping
2012-07-18 16:40:39.553 [IPClusterStart] Process '/usr/bin/python' stopped:
 {'pid': 16130, 'exit_code': 0}
2012-07-18 16:40:42.458 [IPClusterStart] Removing pid file:
 /home/wimmer/.ipython/profile_ssh/pid/ipcluster.pid


_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ipcluster in ssh mode in ipython 0.13

MinRK
Holy crap, this is a one-character typo preventing the default value for remote_profile_dir from being derived from the given profile_dir, leaving it as an empty string.  Sorry about that!

You can workaround this by setting the remote profile dir explicitly with:

c.SSHEngineSetLauncher.remote_profile_dir='.ipython/profile_ssh'

(or similar)

PR #2163 (one-character patch) should fix it.  Can you confirm that?

-MinRK

On Wed, Jul 18, 2012 at 9:47 AM, M. Wimmer <[hidden email]> wrote:
> Dear all,
>
> Having freshly installed ipython 0.13 I tried to get an ipython parallel cluster
> running via ssh. For this I followed precisely the documentation and did
>
>  ipython profile create --parallel --profile=ssh
>
> and modified, according to the docs, in ipcluster_config.py the lines
>
>  c.IPClusterEngines.engine_launcher_class = 'SSHEngineSetLauncher'
>
>  c.SSHEngineSetLauncher.engines = {'xxx049': 1}
>
> Trying to start a cluster then misarebly fails
>
>  ipcluster start --profile=ssh --debug
>
> and I've added the debug output below. The failed `scp`-commands I can avoid by
> writing in ipcluster_config.py
>
>  c.SSHLauncher.to_send = []
>  c.SSHLauncher.to_fetch = []
>
> As you can see from the logfile below, the engines are not started because of
> arguments (""--profile ...") being added to the python command. I can change the
> engine_cmd to `ipengine` but then I ran into other troubles, so I wanted to
> discuss this step by step.
>
> What am I doing wrong? What should be added to the docs?
>
> - Michael
>
> Logfile:
> ---------------------------------------------------------
> 2012-07-18 16:40:33,001.001 [IPClusterStart] Config changed:
> 2012-07-18 16:40:33,001.001 [IPClusterStart] {'Application': {'log_level': 10},
>  'BaseIPythonApplication': {'profile': u'ssh'}}
> 2012-07-18 16:40:33,003.003 [IPClusterStart] Using existing profile dir:
>  u'/home/wimmer/.ipython/profile_ssh'
> 2012-07-18 16:40:33,003.003 [IPClusterStart] Searching path [u'/home/wimmer',
>  u'/home/wimmer/.ipython/profile_ssh'] for config files
> 2012-07-18 16:40:33,003.003 [IPClusterStart] Attempting to load config file:
>  ipython_config.py
> 2012-07-18 16:40:33,005.005 [IPClusterStart] Loaded config file:
>  /home/wimmer/.ipython/profile_ssh/ipython_config.py
> 2012-07-18 16:40:33,005.005 [IPClusterStart] Attempting to load config file:
>  ipcluster_config.py
> 2012-07-18 16:40:33,007.007 [IPClusterStart] Loaded config file:
>  /home/wimmer/.ipython/profile_ssh/ipcluster_config.py
> 2012-07-18 16:40:33,009.009 [IPClusterStart] Config changed:
> 2012-07-18 16:40:33,009.009 [IPClusterStart] {'Application': {'log_level': 10},
>  'BaseIPythonApplication': {'profile': u'ssh'}, 'IPClusterEngines':
>  {'engine_launcher_class': 'SSHEngineSetLauncher'}, 'SSHEngineSetLauncher':
>  {'engines': {'xxx049': 2}}, 'ProfileDir': {}}
> 2012-07-18 16:40:33.012 [IPClusterStart] Starting ipcluster with [daemon=False]
> 2012-07-18 16:40:33.012 [IPClusterStart] Creating pid file:
>  /home/wimmer/.ipython/profile_ssh/pid/ipcluster.pid
> 2012-07-18 16:40:33.013 [IPClusterStart] Starting Controller with
>  LocalControllerLauncher
> 2012-07-18 16:40:33.013 [IPClusterStart] Starting LocalControllerLauncher:
>  ['/usr/bin/python', '-c', 'from IPython.parallel.apps.ipcontrollerapp import
>  launch_new_instance; launch_new_instance()', '--profile-dir',
>  u'/home/wimmer/.ipython/profile_ssh', '--cluster-id', u'', '--log-to-file',
>  '--log-level=20']
> 2012-07-18 16:40:33.020 [IPClusterStart] Process '/usr/bin/python' started:
>  16130
> 2012-07-18 16:40:33.747 [IPClusterStart] 2012-07-18 16:40:33,746.746
>  [IPControllerApp] Using existing profile dir: u'/home/wimmer/.ipython
>  /profile_ssh'
> 2012-07-18 16:40:34.012 [IPClusterStart] Starting 2 Engines with
>  SSHEngineSetLauncher
> 2012-07-18 16:40:34.017 [IPClusterStart] waiting for
>  /home/wimmer/.ipython/profile_ssh/security/ipcontroller-client.json
> 2012-07-18 16:40:35.021 [IPClusterStart] sending
>  /home/wimmer/.ipython/profile_ssh/security/ipcontroller-client.json to
>  xxx049:security/ipcontroller-client.json
>  scp: security/ipcontroller-client.json: No such file or directory
> 2012-07-18 16:40:36.792 [IPClusterStart] sending
>  /home/wimmer/.ipython/profile_ssh/security/ipcontroller-engine.json to
>  xxx049:security/ipcontroller-engine.json
>  scp: security/ipcontroller-engine.json: No such file or directory
> 2012-07-18 16:40:37.957 [IPClusterStart] Starting SSHEngineLauncher: ['ssh',
>  '-tt', u'xxx049', '/usr/bin/python', '-c', 'from
>  IPython.parallel.apps.ipengineapp import launch_new_instance;
>  launch_new_instance()', '--profile-dir', u'', '--log-to-file',
>  '--log-level=20']
> 2012-07-18 16:40:37.969 [IPClusterStart] Process 'ssh' started: 16235
> 2012-07-18 16:40:38.075 [IPClusterStart] Starting SSHEngineLauncher: ['ssh',
>  '-tt', u'xxx049', '/usr/bin/python', '-c', 'from
>  IPython.parallel.apps.ipengineapp import launch_new_instance;
>  launch_new_instance()', '--profile-dir', u'', '--log-to-file',
>  '--log-level=20']
> 2012-07-18 16:40:38.085 [IPClusterStart] Process 'ssh' started: 16237
> 2012-07-18 16:40:38.086 [IPClusterStart] Process 'engine set' started: [None,
>  None]
> 2012-07-18 16:40:38.086 [IPClusterStart] 2012-07-18 16:40:34.055 [scheduler]
>  Scheduler started [leastload]
> 2012-07-18 16:40:39.249 [IPClusterStart] bash: -c: line 0: syntax error near
>  unexpected token `--profile-dir'
> 2012-07-18 16:40:39.249 [IPClusterStart] Connection to xxx049 closed.
> 2012-07-18 16:40:39.250 [IPClusterStart] bash: -c: line 0: `/usr/bin/python -c
>  from IPython.parallel.apps.ipengineapp import launch_new_instance;
>  launch_new_instance() --profile-dir  --log-to-file --log-level=20'
> 2012-07-18 16:40:39.273 [IPClusterStart] Process 'ssh' stopped: {'pid': 16235,
>  'exit_code': 1}
> 2012-07-18 16:40:39.451 [IPClusterStart] bash: -c: line 0: syntax error near
>  unexpected token `--profile-dir'
> 2012-07-18 16:40:39.451 [IPClusterStart] Connection to xxx049 closed.
> 2012-07-18 16:40:39.452 [IPClusterStart] bash: -c: line 0: `/usr/bin/python -c
>  from IPython.parallel.apps.ipengineapp import launch_new_instance;
>  launch_new_instance() --profile-dir  --log-to-file --log-level=20'
> 2012-07-18 16:40:39.455 [IPClusterStart] Process 'ssh' stopped: {'pid': 16237,
>  'exit_code': 1}
> 2012-07-18 16:40:39.455 [IPClusterStart] Process 'engine set' stopped:
>  {'xxx049/0': {'pid': 16235, 'exit_code': 1}, 'xxx049/1': {'pid': 16237,
>  'exit_code': 1}}
> 2012-07-18 16:40:39.455 [IPClusterStart]
>             Engines shutdown early, they probably failed to connect.
>
>             Check the engine log files for output.
>
>             If your controller and engines are not on the same machine, you probably
>             have to instruct the controller to listen on an interface other than
> localhost.
>
>             You can set this by adding "--ip='*'" to your
> ControllerLauncher.controller_args.
>
>             Be sure to read our security docs before instructing your controller
> to listen on
>             a public interface.
>
> 2012-07-18 16:40:39.456 [IPClusterStart] IPython cluster: stopping
> 2012-07-18 16:40:39.553 [IPClusterStart] Process '/usr/bin/python' stopped:
>  {'pid': 16130, 'exit_code': 0}
> 2012-07-18 16:40:42.458 [IPClusterStart] Removing pid file:
>  /home/wimmer/.ipython/profile_ssh/pid/ipcluster.pid
>
>
> _______________________________________________
> IPython-User mailing list
> [hidden email]
> http://mail.scipy.org/mailman/listinfo/ipython-user


_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ipcluster in ssh mode in ipython 0.13

M. Wimmer
Thanks for the quick reply!

> (or similar)PR #2163 (one-character patch) should fix it.  Can you confirm
> that?

I tried this, and it certainly fixes the problem with the profile directory
(i.e. the scp-commands work now).

However I have still problems with the engines not being launched due to a
syntax error (only the relevant part, and I add -vvv to ssh):

2012-07-19 10:29:34.700 [IPClusterStart] debug1: Sending command:
 /usr/bin/python -c from IPython.parallel.apps.ipengineapp import
 launch_new_instance; launch_new_instance() --profile-dir
 .ipython/profile_ssh --log-to-file --log-level=20
2012-07-19 10:29:34.763 [IPClusterStart] bash: -c: line 0: syntax error near
 unexpected token `--profile-dir'
2012-07-19 10:29:34.764 [IPClusterStart] bash: -c: line 0: `/usr/bin/python
 -c from IPython.parallel.apps.ipengineapp import launch_new_instance;  
 launch_new_instance() --profile-dir .ipython/profile_ssh --log-to-file
 --log-level=20'

It seems to me that there are some quotation marks missing around `from ...
launch_new_instance()`? (I have to admit that quoting always confuses me)

Indeed, if I add to ipcluster_config.py

 c.SSHEngineSetLauncher.engine_cmd = ['/usr/bin/python', '-c', '\"from
  IPython.parallel.apps.ipengineapp import launch_new_instance;
  launch_new_instance()\"']

the syntax error is solved.

Unfortunately, the cluster still does not start, as the engines simply stop
within two seconds or so. I'm still investigating this, but in any case it
would be nice if you could have a look into what I wrote above.

All the best,

Michael


_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ipcluster in ssh mode in ipython 0.13

Robert Nishihara
I am also interested in the answer to this question.

I think I am running into the same problem (except I am using SGE) when I use IPython 0.13. It works in IPython 0.12. 

On Thu, Jul 19, 2012 at 1:39 AM, M. Wimmer <[hidden email]> wrote:
Thanks for the quick reply!

> (or similar)PR #2163 (one-character patch) should fix it.  Can you confirm
> that?

I tried this, and it certainly fixes the problem with the profile directory
(i.e. the scp-commands work now).

However I have still problems with the engines not being launched due to a
syntax error (only the relevant part, and I add -vvv to ssh):

2012-07-19 10:29:34.700 [IPClusterStart] debug1: Sending command:
 /usr/bin/python -c from IPython.parallel.apps.ipengineapp import
 launch_new_instance; launch_new_instance() --profile-dir
 .ipython/profile_ssh --log-to-file --log-level=20
2012-07-19 10:29:34.763 [IPClusterStart] bash: -c: line 0: syntax error near
 unexpected token `--profile-dir'
2012-07-19 10:29:34.764 [IPClusterStart] bash: -c: line 0: `/usr/bin/python
 -c from IPython.parallel.apps.ipengineapp import launch_new_instance;
 launch_new_instance() --profile-dir .ipython/profile_ssh --log-to-file
 --log-level=20'

It seems to me that there are some quotation marks missing around `from ...
launch_new_instance()`? (I have to admit that quoting always confuses me)

Indeed, if I add to ipcluster_config.py

 c.SSHEngineSetLauncher.engine_cmd = ['/usr/bin/python', '-c', '\"from
  IPython.parallel.apps.ipengineapp import launch_new_instance;
  launch_new_instance()\"']

the syntax error is solved.

Unfortunately, the cluster still does not start, as the engines simply stop
within two seconds or so. I'm still investigating this, but in any case it
would be nice if you could have a look into what I wrote above.

All the best,

Michael


_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user


_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ipcluster in ssh mode in ipython 0.13

Robert Nishihara
<http://lists.ipython.scipy.org/pipermail/ipython-user/2012-July/010619.html> for reference.

On Tue, Jul 24, 2012 at 5:42 PM, Robert Nishihara <[hidden email]> wrote:
I am also interested in the answer to this question.

I think I am running into the same problem (except I am using SGE) when I use IPython 0.13. It works in IPython 0.12. 


On Thu, Jul 19, 2012 at 1:39 AM, M. Wimmer <[hidden email]> wrote:
Thanks for the quick reply!

> (or similar)PR #2163 (one-character patch) should fix it.  Can you confirm
> that?

I tried this, and it certainly fixes the problem with the profile directory
(i.e. the scp-commands work now).

However I have still problems with the engines not being launched due to a
syntax error (only the relevant part, and I add -vvv to ssh):

2012-07-19 10:29:34.700 [IPClusterStart] debug1: Sending command:
 /usr/bin/python -c from IPython.parallel.apps.ipengineapp import
 launch_new_instance; launch_new_instance() --profile-dir
 .ipython/profile_ssh --log-to-file --log-level=20
2012-07-19 10:29:34.763 [IPClusterStart] bash: -c: line 0: syntax error near
 unexpected token `--profile-dir'
2012-07-19 10:29:34.764 [IPClusterStart] bash: -c: line 0: `/usr/bin/python
 -c from IPython.parallel.apps.ipengineapp import launch_new_instance;
 launch_new_instance() --profile-dir .ipython/profile_ssh --log-to-file
 --log-level=20'

It seems to me that there are some quotation marks missing around `from ...
launch_new_instance()`? (I have to admit that quoting always confuses me)

Indeed, if I add to ipcluster_config.py

 c.SSHEngineSetLauncher.engine_cmd = ['/usr/bin/python', '-c', '\"from
  IPython.parallel.apps.ipengineapp import launch_new_instance;
  launch_new_instance()\"']

the syntax error is solved.

Unfortunately, the cluster still does not start, as the engines simply stop
within two seconds or so. I'm still investigating this, but in any case it
would be nice if you could have a look into what I wrote above.

All the best,

Michael


_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user



_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ipcluster in ssh mode in ipython 0.13

MinRK
The SGE issue is fixed in master, and a workaround is described in Issue #2171.

On Tue, Jul 24, 2012 at 5:44 PM, Robert Nishihara <[hidden email]> wrote:
> <http://lists.ipython.scipy.org/pipermail/ipython-user/2012-July/010619.html>
> for reference.
>
>
> On Tue, Jul 24, 2012 at 5:42 PM, Robert Nishihara
> <[hidden email]> wrote:
>>
>> I am also interested in the answer to this question.
>>
>> I think I am running into the same problem (except I am using SGE) when I
>> use IPython 0.13. It works in IPython 0.12.
>>
>>
>> On Thu, Jul 19, 2012 at 1:39 AM, M. Wimmer <[hidden email]> wrote:
>>>
>>> Thanks for the quick reply!
>>>
>>> > (or similar)PR #2163 (one-character patch) should fix it.  Can you
>>> > confirm
>>> > that?
>>>
>>> I tried this, and it certainly fixes the problem with the profile
>>> directory
>>> (i.e. the scp-commands work now).
>>>
>>> However I have still problems with the engines not being launched due to
>>> a
>>> syntax error (only the relevant part, and I add -vvv to ssh):
>>>
>>> 2012-07-19 10:29:34.700 [IPClusterStart] debug1: Sending command:
>>>  /usr/bin/python -c from IPython.parallel.apps.ipengineapp import
>>>  launch_new_instance; launch_new_instance() --profile-dir
>>>  .ipython/profile_ssh --log-to-file --log-level=20
>>> 2012-07-19 10:29:34.763 [IPClusterStart] bash: -c: line 0: syntax error
>>> near
>>>  unexpected token `--profile-dir'
>>> 2012-07-19 10:29:34.764 [IPClusterStart] bash: -c: line 0:
>>> `/usr/bin/python
>>>  -c from IPython.parallel.apps.ipengineapp import launch_new_instance;
>>>  launch_new_instance() --profile-dir .ipython/profile_ssh --log-to-file
>>>  --log-level=20'
>>>
>>> It seems to me that there are some quotation marks missing around `from
>>> ...
>>> launch_new_instance()`? (I have to admit that quoting always confuses me)
>>>
>>> Indeed, if I add to ipcluster_config.py
>>>
>>>  c.SSHEngineSetLauncher.engine_cmd = ['/usr/bin/python', '-c', '\"from
>>>   IPython.parallel.apps.ipengineapp import launch_new_instance;
>>>   launch_new_instance()\"']
>>>
>>> the syntax error is solved.
>>>
>>> Unfortunately, the cluster still does not start, as the engines simply
>>> stop
>>> within two seconds or so. I'm still investigating this, but in any case
>>> it
>>> would be nice if you could have a look into what I wrote above.
>>>
>>> All the best,
>>>
>>> Michael
>>>
>>>
>>> _______________________________________________
>>> IPython-User mailing list
>>> [hidden email]
>>> http://mail.scipy.org/mailman/listinfo/ipython-user
>>
>>
>
>
> _______________________________________________
> IPython-User mailing list
> [hidden email]
> http://mail.scipy.org/mailman/listinfo/ipython-user
>


_______________________________________________
IPython-User mailing list
[hidden email]
http://mail.scipy.org/mailman/listinfo/ipython-user
Loading...