From owner-freebsd-questions Wed Aug 2 13:43:19 1995 Return-Path: questions-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.11/8.6.6) id NAA29007 for questions-outgoing; Wed, 2 Aug 1995 13:43:19 -0700 Received: from nero.uucp (dialup-pkr-1-5.network.umr.edu [131.151.253.6]) by freefall.cdrom.com (8.6.11/8.6.6) with SMTP id NAA28982 for ; Wed, 2 Aug 1995 13:43:07 -0700 Received: by nero.uucp (Smail3.1.29.1 #1) id m0sdkah-0004JaC; Wed, 2 Aug 95 15:40 CDT Message-Id: From: serges@umr.edu Subject: Re: 2.0.5 Eager to go into swap To: freebsd-questions@freefall.cdrom.com Date: Wed, 2 Aug 1995 15:40:35 -0500 (CDT) In-Reply-To: <199508021755.KAA19004@freefall.cdrom.com> from "John Dyson" at Aug 2, 95 10:55:34 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 2832 Sender: questions-owner@FreeBSD.org Precedence: bulk > > > > I have a 486dx2/66 with 32 Megs of ram, and a 32 meg swap partition. > > When I was using 2.0 (CD) i rarely ever went into swap. I could make a > > port, and run netscape under X and still not hit swap. > > > > Now with 2.0.5, it seems very eager to start swaping. In fact, with the > > above, it goes into 45% or more swap. This sucks to say the least. My > > machine is trashing big time when this happens. Is anyone familiar with > > this? Is there a fix? I know swaping was changed in 2.0.5, but I heard > > it was for the better, not the worse..... > > > > Any suggestions, comments, are most welcome. > > > > Since both the kernel and the user-land has changed, it is difficult for > a user to be able to determine where the problem is. The kernel itself > does swap better. There are some very limited cases where it *might* be > worse on 2.0.5, but I don't think that you are exercising those mechanisms. Well I dont know what those "mechanisms" are and I (unlike the original poster) cant compare 2.0.5 with a previous version of FreeBSD, but I can say that I experience excessive swapping on my system. I have a 486DX50 with 20 megs of core and ~20 megs of swap space. With a minimal X desktop with Netscape and 2 xterms running I can *easily* exhaust the virtual memory on my system! This is ofcourse, after running Netscape (or xv) for a long time (> 1hour continuous use). I usually have to kill the server and restart things. I can understand that the server grows, thanks to memory leaks in X11 itself, but I dont understand why the 2.0.5 system begins swapping *before* the core resources are even 50% in use. Coming from a SVR4 background on Intel machines, I find it a hassle to watch as all of my xterms slowly redraw themselves as they are swapped back in to core - after I briefly switch back from another application. And I dont think its solely the fault of the clients and their real or virtual sizes. I feels more like a scheduling design at fault here. But Im speaking as a user (of Unices on many different platforms), and not a kernel programmer. Serge serges@umr.edu -------------- > Probably the reason that you are seeing more paging is because the X server > appears to grow bigger than it used to. This can be because of changes > in X clients, changes in the system (user-land) malloc, or changes in the > X-server itself. > > The current resident size of any given process can be obtained by using > the ps command and looking at the RSS field. The RSS field only accounts > for the pages that are currently mapped into the process and ignores > any aspect of sharing or pages on disk. > > The VSZ field might be educational for you... It *does* get big for > XFree V3.1.1 under 2.0.5 (and other OSes like Linux also). > > John > dyson@freebsd.org > >