Thanks for your answers, this will probably solve all my problems. And
writing to a file sounds good, since I can always read it's tail and
do whatever I want.
Yann
On 17 sep, 22:03, John Voight <[EMAIL PROTECTED]> wrote:
> Hello!
>
> It is a pity that Yi has moved on (at least for the moment), but it
> doesn't surprise me: he's amazing and surely working on fabulous and
> interesting things.
>
> I did and do use dsage quite extensively--it's been essential for me
> in enumerating number fields. I also have access to a cluster of
> computers, and instead of queuing various jobs I just run them as
> dsage workers whenever there is backflow. This works really well.
>
> It took some debugging to get it all to work satisfactorally, and I
> still don't know terribly much about what's under the hood so to
> speak, but I'm happy to help.
>
> > 1/ Why isn't there a clear, publicized, illustrated description of how to
> > use dsage ? I managed to make it work, but only after googling hard.
>
> +1
>
> > 2/ How can I send a job to a worker that will output intermediate values ?
> > I mean, say the job sent to a particular worker computes some value, and
> > that it takes 100 iterations, how can I output temporary values every 10
> > iterations and have the server report those intermediate values ?
>
> Do you need to know those intermediate values really intermediately,
> or just in the end? If it's the former, then you should probably
> break up your task into computing 10 iterations and report each of
> those back. If it's the latter, then just return the intermediate
> values as part of the output. If you're not using them for anything,
> you can always also just write to a file...
>
> > 3/ I noticed that workers can connect any time, really, and receive jobs
> > even if they connect to the server only after the server started some
> > sequence of jobs, which is cool. But I also noticed that if a worker gets
> > killed, then its job gets lost. Isn't it possible for the server to check
> > if a worker is alive, every once in a while, and if not requeue its job ?
>
> I don't know about checking workers, but this happens to me
> frequently. I use the failures attribute of a job to see if it has
> completed or not and requeue that way. Dead workers don't get new
> jobs.
>
> > 4/ What is the function I can use to check which worker did what, and if
> > it's alive, and what job got interrupted.
>
> Not sure this has been implemented. I was only ever interested in
> total computing time, which you can recover from the cpu_time
> attribute. What would you do with this information?
>
> > 5/ What test can I apply to a dsage job to see if it's finished ? Say a
> > job outputs a list, and I want to plot it, can I say something like "If
> > there is some output, plot it, otherwise wait." ?
>
> Use the get_job() function to retrieve the output and then the status
> attribute ('new', 'processing', or 'completed') to check if it is
> done. If you just want to wait until a job is finished, you can use
> the wait function on a job--or you can also repeatedly check and use
> the time.sleep() function.
>
> > 6/ If you have any notes, drafts, illustrating some of dsage
> > functionalities, I'd be more than happy to check them out.
>
> You can check out the functions in
> sage.rings.number_field.totallyreal_dsage; I'm not saying they're
> optimal, but they (mostly) work, so might be of some inspiration.
>
> Let me know if I can be of any other help.
>
> Yours,
>
> John Voight
> Assistant Professor of Mathematics
> University of Vermont
> [EMAIL PROTECTED]
> [EMAIL PROTECTED]://www.cems.uvm.edu/~voight/
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-support@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at
http://groups.google.com/group/sage-support
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
