Re: [scipion-users] Scipion 3.0 Issue with Commands Waiting for Eachother

[Pablo C. <pc...@cn...>]

Thanks Colin, see anwser in line.

On 1/10/20 23:49, Colin Deniston wrote:
>
> Hello,
>
> First off, I’m new to Scipion and have been loving the flexibility, 
> streaming, monitoring, and reporting abilities of the program! 
> However, I’m currently trying to run a custom built workflow in 
> Scipion 3.0.3 and I’m running into a few issues.
>
> 1) I don’t see in the actual .json.template file a line to designate 
> what ID to wait for under each protocol (I want each job to wait for 
> the prior job to completely finish before running). Is this something 
> I can add? If so, what is the command I should use?
>
Certainly something to Implement. Jose Miguel is right about the ids 
potentially been changes, but I think we can update the "wait for" 
attribute too (and export it/import it). I've created and issue 
<https://github.com/scipion-em/scipion-pyworkflow/issues/151> and will 
look at it:
>
> 2) When I try and circumvent that by simply typing the ID in the GUI 
> once the workflow is loaded (in the “wait for” box of each protocol), 
> nothing seems to wait. They all run while the previous job is still 
> going. Am I doing something wrong? I’ve been entering them as just the 
> ID number like “432” but should they be entered differently? Also, I 
> see multiple IDs for the same job/protocol under different menus, not 
> sure if that is a bug or not or whether it is maybe related to the 
> issue of jobs not waiting.
>
The expected behavior is all should be "scheduled" (orange). This does 
not mean they are running, but they are actively checking is they can 
start once all validation passes (wait for should be part of the validation)

Do the box with the "wait for" turns orange (scheduled,this would be 
correct) or green (running, thus not waiting)?

> Any feedback would be helpful, my end goal is just to make a 
> reproducible workflow that can be loaded and run on multiple 
> Appoferritin calibration datasets all the way to 3D refinement without 
> any user intervention. However, at the moment one job will try to run 
> too soon and fail and then lead to a cascade of downstream jobs 
> failing. Thanks!
>
> *Colin Deniston*
> Principal Scientist I
>
> *GNF BIOTHERAPEUTICS & BIOTECHNOLOGY / GNF*
>
> Novartis Institutes for Biomed.Research, Inc.
>
> 10675 John Jay Hopkins Drive
>
> GNF Business
>
> San Diego, CA 92121
>
> T +1 (858) 332-4750
>
> cde...@gn... <mailto:cde...@gn...>
>
> https://saps-app.prd.nibr.novartis.net/esh/novartislogo.png 
> <https://www.novartis.com/>
>
>
>
> _______________________________________________
> scipion-users mailing list
> sci...@li...
> https://lists.sourceforge.net/lists/listinfo/scipion-users
-- 
Pablo Conesa - *Madrid Scipion <http://scipion.i2pc.es> team*