From: Roger Pack <rogerpack2005@...>
Date: 2008-08-16T05:37:05+09:00
Subject: Re: ruby wish-list

>> Oh, I wasn't understanding your idea, which does seem to make sense now.
>> There's the overhead of essentially duplicating the object space of the
>> original process in the fork (since mark will write to memory pages).
>> Maybe it would be worth this cost (esp. on a second processor) for the
>> benefit of not having to stop the main process?
...
> I'll probably hack it up sometime, first version as a straight patch to 
> 1.8.6

Turns out that just increasing the malloc limit to 80MB or so gives a 
nice decrease in GC time[1]--about 80% for me.  So that's an easy stop 
gap till I get this going.  It does tend to use a lot more RAM, though, 
unfortunately.

Question about typical VPN's: I assume that typically the bottleneck is 
typically RAM and not CPU [ex: on 256MB slice], and they have access to 
multiple cores where available, is that right?
Thanks!
-=R
[1] http://github.com/skaes/railsbench/tree/v0.8.3/GCPATCH
-- 
Posted via http://www.ruby-forum.com/.