On Wed, 08 Nov 2006 03:32:53 -0800, Martin Albrecht <[EMAIL PROTECTED]> wrote:
> >> (5) The overhead is minimal -- it really is only 2MB to run a minimal >> Python process. > > That can/should be reduced further either by moving this stuff to C or > the > bash. I thought that too -- but 2MB is pretty minimal, and I'm not convinced it is better to do that in C. Moreover, I imagine running the monitor process remotely on machines over the net, which may have a different architecture, which might not even have SAGE; this will be way easier if it's a Python script since that only requires there be some Python on the remote machine. If it's a C program, you have to compile that program on the remote machine, which is unrealistic in general. By the way, the remote monitor script for running processes interactively via ssh over a network connection will obviously have to be different than the one I wrote, since it can't check if the master SAGE process is still running. They would have to check something else, which I haven't figured out yet. (Ideas?) > Also, I'm under the impression that the whole zombie thing is a > regression. > E.g., whenever I now kill a Gröbner basis calculation using Singular a > Singular process keeps running (insane). I don't remember this happening > circa 2 month ago. Anyway. Not exactly. The killing of subprocesses did get worse with some recent changes; and in fact in the sage-dev version of SAGE (the next version) things will be as they were before (and in fact better). However, there is no possible way that the model from a few weeks ago is sufficient, since e.g., if you "kill -9" the parent SAGE process, there's no way for it to properly clean up subprocesses. With the new monitor model this cleaning up of subprocesses works fine. William --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@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-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~----------~----~----~----~------~----~------~--~---
