Re: [galaxy-dev] Improved DRMAAJobRunner (Peter van Heusden)

2017-07-08 Thread Peter van Heusden 
On Sat, 8 Jul 2017 at 11:56 Matthias Bernt  wrote:

> Hi Peter,
>
> > Code looks interesting. This is for Univa Grid Engine?
>
> In its current state its tailored for the Univa Grid Engine. But the drmaa
> library code that is based on job_info() and wait() should run on all grid
> engines.
> For the command line based code changes might be necessary for the
> different grid engines. (Currently there is only a small bug, because
> wait() is currently
> called twice for finished jobs: in the repeated polling and then in the
> final check. The second call won't work which causes a fallback to qacct. I
> will fix this
> soon.)
> I guess that the output of qstat and qacct will be different. But I think
> this can be configured one just needs a way to get the info which grid
> engine is running.
>
> > In terms of the upload jobs, are those not designed to be run as 'local'
> jobs and not with the 'real user' setting?
>
> Sounds reasonable, but this is not what is happening on my installation of
> galaxy. Any idea where I could start to find the problem.
>
> I'm not an expert on this, but what does your job_conf.xml look like?

Peter

> Best,
> Matthias ___
> Please keep all replies on the list by using "reply all"
> in your mail client.  To manage your subscriptions to this
> and other Galaxy lists, please use the interface at:
>   https://lists.galaxyproject.org/
>
> To search Galaxy mailing lists use the unified search at:
>   http://galaxyproject.org/search/
___
Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:
  https://lists.galaxyproject.org/

To search Galaxy mailing lists use the unified search at:
  http://galaxyproject.org/search/

Re: [galaxy-dev] Improved DRMAAJobRunner (Peter van Heusden)

2017-07-08 Thread Matthias Bernt 
Hi Peter, 

> Code looks interesting. This is for Univa Grid Engine?

In its current state its tailored for the Univa Grid Engine. But the drmaa 
library code that is based on job_info() and wait() should run on all grid 
engines. 
For the command line based code changes might be necessary for the different 
grid engines. (Currently there is only a small bug, because wait() is currently 
called twice for finished jobs: in the repeated polling and then in the final 
check. The second call won't work which causes a fallback to qacct. I will fix 
this 
soon.) 
I guess that the output of qstat and qacct will be different. But I think this 
can be configured one just needs a way to get the info which grid engine is 
running. 

> In terms of the upload jobs, are those not designed to be run as 'local' jobs 
> and not with the 'real user' setting?

Sounds reasonable, but this is not what is happening on my installation of 
galaxy. Any idea where I could start to find the problem. 

Best, 
Matthias
___
Please keep all replies on the list by using "reply all"
in your mail client.  To manage your subscriptions to this
and other Galaxy lists, please use the interface at:
  https://lists.galaxyproject.org/

To search Galaxy mailing lists use the unified search at:
  http://galaxyproject.org/search/

Re: [galaxy-dev] ustacks - bad interpreter

2017-07-08 Thread Björn Grüning 
Hi,

please also make sure you have a recent conda version. This 80 character
length smalls a lot like an old conda restriction.

Cheers,
Bjoern

Am 07.07.2017 um 18:09 schrieb David Meltzer:
> Good afternoon,
> 
>  
> 
> Thank you for that! I will look in to it!
> 
>  
> 
> Best Regards,
> 
> David Jacob Meltzer
> 
>  
> 
> *From: *Nate Coraor 
> *Date: *Friday, July 7, 2017 at 5:02 PM
> *To: *David Meltzer 
> *Cc: *Yvan Le Bras ,
> "galaxy-dev@lists.galaxyproject.org" 
> *Subject: *Re: [galaxy-dev] ustacks - bad interpreter
> 
>  
> 
> Hi David,
> 
>  
> 
> You may want to have a look at the first line of:
> 
>  
> 
> /home/big_galaxy/galaxy/tool_dependencies/_conda/envs/mulled-v1-8e6d35eac718a13db6458b9361ca87c93be9feed656f9938ebcc435642f3a0a3/bin/stacks_summary.py
> 
>  
> 
> The shell is claiming it is:
> 
>  
> 
> #!/home/big_galaxy/galaxy/tool_dependencies/_conda/envs/mulled-v1-8e6d35eac718a1
> 
>  
> 
> When it should be:
> 
>  
> 
> #!/home/big_galaxy/galaxy/tool_dependencies/_conda/envs/mulled-v1-8e6d35eac718a13db6458b9361ca87c93be9feed656f9938ebcc435642f3a0a3/bin/python
> 
>  
> 
> However, it could be that the line is correct and the output was
> truncated, or the shell has a length limit for the shebang (although the
> truncated string is only 79 characters, I'd be surprised if the limit
> for your shell was that short).
> 
>  
> 
> --nate
> 
>  
> 
> On Fri, Jul 7, 2017 at 8:21 AM, David Meltzer
> > wrote:
> 
> Good afternoon,
> 
>  
> 
> I will do my best!
> 
> 
> Best Regards,
> David
> 
>  
> 
> *From: *Yvan Le Bras  >
> *Date: *Friday, July 7, 2017 at 1:06 PM
> 
> 
> *To: *David Meltzer  >,
> "galaxy-dev@lists.galaxyproject.org
> "
>  >
> *Subject: *RE : [galaxy-dev] ustacks - bad interpreter
> 
>  
> 
> No problem, I was not writing my question clearly...
> 
>  
> 
> I suspect a pb with the stacks_summary related script... Can you
> investigate it, I unfortunately can't work on it for now. .. Maybe
> ""just"" a missing python interpreter invocation somewhere. .. when
> executing stacks_summary. ...
> 
> 
> 
>  Message d'origine 
> De : David Meltzer  >
> Date : 07/07/2017 13:47 (GMT+01:00)
> À : Yvan Le Bras  >, galaxy-dev@lists.galaxyproject.org
> 
> Objet : Re: [galaxy-dev] ustacks - bad interpreter
> 
>  
> 
> Good afternoon,
> 
>  
> 
> I apologize. I misunderstood the question. Yes, stacks_summary has
> been installed correctly as have stacks and velvet.
> 
>  
> 
> Additionally, the tool is outputting three items into the history -
> 
>  
> 
> Summary from Stacks – red (error)
> 
> Ustacks.log – red (error)
> 
> Stacks from Data – green (successful)
> 
>  
> 
> Best Regards,
> 
> David Jacob Meltzer
> 
>  
> 
> *From: *galaxy-dev  > on behalf of
> David Meltzer  >
> *Date: *Friday, July 7, 2017 at 12:31 PM
> *To: *Yvan Le Bras  >, "galaxy-dev@lists.galaxyproject.org
> "
>  >
> *Subject: *Re: [galaxy-dev] ustacks - bad interpreter
> 
>  
> 
> Good afternoon,
> 
>  
> 
> I am sorry. I don’t see a tool called stacks_summary installed on my
> system or available on the toolshed.
> 
>  
> 
> Best Regards,
> 
> David
> 
>  
> 
> *From: *Yvan Le Bras  >
> *Date: *Friday, July 7, 2017 at 12:08 PM
> *To: *David Meltzer  >,
> "galaxy-dev@lists.galaxyproject.org
> "
>  >
> *Subject: *RE : [galaxy-dev] ustacks - bad interpreter
> 
>  
> 
> Hi David,
> 
>  
> 
> A pb with stacks_summary apparently... and ustacks is working well
> apparently... Can you check you have stacks_summary well installed ?
> I think it is but there is a