[IPython-dev] Before a patch for LSF support

Matthieu Brucher matthieu.brucher@gmail....
Thu Aug 13 03:19:57 CDT 2009


2009/8/13 Brian Granger <ellisonbg.net@gmail.com>:
>
>> OK, I have some improvements here:
>> - I can connect to the controller, but I have first to copy the furl:
>> $HOME on the headnode is not available on the computenode
>> - with ipcontroller-engine.furl, it seems to be working.
>
> Great, that is definitely progress!
>
>>
>> So the only issue is that I need to copy this furl file to the node,
>> as they can't access the file. I may have had some furl files left
>> inside that folder, which lead to the issue at hand. This should be
>> done with --engine-furl-file=... I suppose? Is there a way of exposing
>> it inside ipcluster, with kernel_config perhaps?
>
> Yes, stale furl files can definitely cause problems if you are not using
> persistent furl files.  Currently, ipcluster assumes that you will be using
> a shared file system that all the engines and controller can read in the
> same location.  Thus, we don't yet expose options like this to ipcluster.
> BUT, we are in the middle of completely refactoring IPython's config
> system.  Part of that work will be to make ipcluster more configurable.
> This should happen sometime in the next 2 months.  In the meantime if you
> have patches to add such an option to ipcluster, that would be fine.  Keep
> us posted.

Excellent ! I saw a mail of Fernando on this. I will need some time to
help you with this.

Meanwhile, I will try to use ipcluster from a node to another LSF
node. I have still some issues with the IP address: when I'm starting
ipcluster from a compute node, the furl points to 127.0.0.1 instead of
the name of the host or at least its public IP address.

cheers,

Matthieu
-- 
Information System Engineer, Ph.D.
Website: http://matthieu-brucher.developpez.com/
Blogs: http://matt.eifelle.com and http://blog.developpez.com/?blog=92
LinkedIn: http://www.linkedin.com/in/matthieubrucher


More information about the IPython-dev mailing list