On May 3, 6:48 pm, "brandon.bar...@gmail.com"
<brandon.bar...@gmail.com> wrote:
> Hello,
Hi,
> I'm very happy about the improved Solaris support. Until recently I
> was running sage in a linux zone on my workstation, but unfortunately
> modern linux version support for branded zones is lagging. I confess
> I've also been using sagenb.org a bit (which will be running on a
> T2000 soon?)
Probably, but the GMP/MPIR performance on the T2000 is pathetic per
"thread" and in effect only 8 or 16 out of 64 threads can use the
integer or FP units at a time, so scaling out kind of sucks, too.
> I've tried to run the notebook in 3 different
> environments (all by simply calling notebook() from the sage shell).
>
> 1: Updated OpenSolaris 2008.11 (snv_111a)
>
> 2: SXCE global zone (snv_109)
>
> 3: SXCE sparse root zone (same machine as (2))
>
> On all of these, the startup of the notebook seems to hang for several
> hours at this point:
If the notebook takes forever to initialize this indicates a problem
with the source of randomness. I know too little about your setup to
determine if it is the system not providing enough entropy to the zone
or what else could be interfering here. libgcrypt (which is used to
create the the keys) used to have a bug that caused it to use way more
entropy then required and this caused a very long hang on Solaris 10.
We have since updated and at least on plain vanilla Solaris 10 (i.e.
no zone) the notebook start is instant. Any chance this gives you a
hint what is wrong? Does the problem also happen outside the zone?
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to
sage-devel-unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
