From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from pigeon.gentoo.org ([208.92.234.80] helo=lists.gentoo.org) by finch.gentoo.org with esmtp (Exim 4.60) (envelope-from ) id 1QPeLn-0000qp-MC for garchives@archives.gentoo.org; Thu, 26 May 2011 17:27:59 +0000 Received: from pigeon.gentoo.org (localhost [127.0.0.1]) by pigeon.gentoo.org (Postfix) with SMTP id 727C11C198; Thu, 26 May 2011 17:26:24 +0000 (UTC) Received: from mx.virtyou.com (mx.virtyou.com [94.23.166.77]) by pigeon.gentoo.org (Postfix) with ESMTP id 302F81C198 for ; Thu, 26 May 2011 17:26:24 +0000 (UTC) Received: from weird.localnet (p5B275544.dip.t-dialin.net [91.39.85.68]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by mx.virtyou.com (Postfix) with ESMTPSA id 34F6B4A82BD for ; Thu, 26 May 2011 19:26:23 +0200 (CEST) To: gentoo-user@lists.gentoo.org Subject: Re: [gentoo-user] Swap performance From: Alex Schuster Date: Thu, 26 May 2011 19:26:20 +0200 Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-Id: Gentoo Linux mail X-BeenThere: gentoo-user@lists.gentoo.org Reply-to: gentoo-user@lists.gentoo.org MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Message-Id: <201105261926.21424.wonko@wonkology.org> X-Archives-Salt: X-Archives-Hash: 3127d4e3b16cbce0861cfc48f62fc30d Volker Armin Hemmann writes: > On Thursday 26 May 2011 00:40:21 Alex Schuster wrote: > > Paul Hartman writes: > > > On Wed, May 25, 2011 at 9:20 AM, Alex Schuster > > > wrote: > > Maybe I should have bought 4G instead of 2G, so I'd have 10G, not 8G. > > I'm not sure if this is recommended these days, to have one memory bank > > with 4G and 3 others with 2G each. > > It is very much not recommended nor wise. > > Sometimes it works great, sometimes it just works, sometimes it burns > down your village, rapes your cattle and steals your dad. Ookay. > > It's growing: 1405m VIRT, 851m RES, 6m SHR. Strange, I did not actually > > use the desktop after I wrote the mail you replied to, currently I'm > > logged in from remote. > > and you know that those numbers are pretty much meaningless? Are they? Well, apparently _something_ in kwin was using hundreds of megabytes. When I was back at the PC, I had to wait three minutes until he password dialog appeared so I could unlock the session. iotop showed the kwin process swapping, and no other activity. Now kwin is at 37M, which is much lower than I thought... whoops, I made a mistake in my previous posting, I got the virtual memory column instead by misusing awk. I also confused kwin with plasma-desktop, which had been really high in the past. > about swap: > > 8gb ram, 24gb swap here. Swap so huge because of historical reasons > (started with one disk with 8gb, now there are three...). But not so > bad, considering all those tempfs mounts that can shoved in there. Yikes. I'd rather use the space for other stuff... no matter how large hard drives are, mine tend to fill up so I am glad for every extra gigabyte I can find. > Of course I am scared about the shitstorm if that ever happens. > > I am using the standard scheduler, no fancy io-scheduling stuff, kernel > 2.6.36.6 and can't complain. I do, a lot. But it seems that the system performs okay now. > Even after a week of uptime I only get 500mb swap. Some cruft still in > memory for some i-dont-know-reasons shoved in the hellhole swap so it > won't get in the way of the more important stuff. Like gwenview. Or vlc. Yeah, that's how it should be. A little swap is okay. But in my case it felt like important stuff was swapped out. Wonko