Discussion related to cado-nfs

[paul zimmermann <Paul.Zimmermann@inria.fr>]

       Hi,

> Hello! First of all I want to say Thankyou for your software. It starts
> working just out of the box, and moreover it can be started on several
> mchines, it's great!

thank you for those kinds words.

> I'm working in Russian antivirus company Dr.Web and for the last 4 years
> I'm fighting with encoders aka cryptolockers which encrypts user's files
> and ask for money to decrypt them. Now we're started to deal with RSA, so
> we need a software to factorize integers, so I have some questions about
> cado-nfs.
> 
> 1) How cado-nfs works on multiple-core systems? For example, I'm planning
> to use 32-CPU-cores servers. Should I specify threads number in
> commandline, or cado-nfs will scale on all the cores?

on a single computer with several cores, the main script (cado-nfs.py) will
automatically detect the number of cores. Alternatively you can specify it
with cado-nfs.py -t <number_of_threads>.

> 2) I want to use cado-nfs in multiple-machine mode with server and slaves.
> What will happens if some of the slaves will be terminated without
> finishing it's work? Will the master reassign it's task to other slaves?

there is some timeout mechanism (see tasks.wutimeout in the file
parameters/factor/params.c90). When the timeout is achieved, the
server will reassign the work units.

> 3) Is it possible to see a statistics in multi-machine mode: how many
> slaves do we have, how many active slaves, how many CPU-cores etc?

this is not yet available.

Best regards,
Paul Zimmermann