Bill Hart wrote: > Hello,
> To possibly add to the sage.math woes yesterday, I was testing a new > memory manager just before the slowdown. It had the unusual property > of allocating very many gigabytes of swap space in small chunks and > very little actual memory. My presumption was that the slowness had > something to do with sage.math trying to recover from the swap space > mess it created for itself after I killed the process that was causing > the problem. > Nope, when I checked there was less than 500MBs of swap used while there were more than 8GB of physical RAM available (even accounting for buffers). You also cannot allocate swap space directly because the allocs are buffered in system ram. The same thing holds for files mmaped into system RAM. The first thing I checked upon investigation was if somebody had pushed the box into swap and that wasn't the case. An 'ls -1 > foo.txt' in the now infamous tmp-directory took in excess of 10 seconds which is pretty bad and that was after an undetermined number of files had been deleted. > Needless to say, the new new memory manager doesn't have this > "feature". > One would hope so :) > Sorry if my tinkering contributed to the problem. > > Bill. > Cheers, Michael <SNIP> --~--~---------~--~----~------------~-------~--~----~ 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/ -~----------~----~----~----~------~----~------~--~---
