Hi Muli, First, 10x for the input. Anyhow, what I'm trying to figure out is why RLIMIT_RSS (or ulimit -m) has no effect at all.
Noam On 10/26/05, Muli Ben-Yehuda <[EMAIL PROTECTED]> wrote: > On Wed, Oct 26, 2005 at 11:20:42AM +0200, Noam Meltzer wrote: > > Hi, > > I'm trying to set a size limit over processes for some system. > > I tried setting 'ulimit -m' (memory size limit), but it had no effect. > > After playing around with ulimit (at the end, i will put the > > configuration in /etc/security/limits.conf) I found that when I > > change: 'ulimit -v' (virtual memory) it can restrict the size of my > > processes. But the problem is that I can't find a direct relation > > between the size I set in the 'ulimit -v' command and the actual size > > i'm allowed to allocate. > > Can anyone explain or direct me to information of what is this > > relation between the two parameters? > > -v - virtual memory size - maximum size of memory the process could > theoretically use. > -m - rss size - maximum size of the resident set - how much *physical* > memory the process could use at any given point in time. > Both are counted in pages. > > With regards to -m and its apparent non-usefullness, setrlimit(2) has > this to say: > > RLIMIT_RSS > Specifies the limit (in pages) of the process's resident set > (the number of virtual pages resident in RAM). This limit only > has effect in Linux 2.4 onwatrds, and there only affects calls > to madvise() specifying MADVISE_WILLNEED. > > If you really need this, I suggest you take a look at the overcommit > patch and the various beancounting patches. > > Cheers, > Muli > -- > Muli Ben-Yehuda > http://www.mulix.org | http://mulix.livejournal.com/ > > ================================================================To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
