Date: Mon, 22 Nov 2004 12:12:52 -0500 From: Charles Swiger <cswiger@mac.com> To: freebsd-questions@freebsd.org Subject: Re: Tracing Disk Activity Message-ID: <BE4BE0F1-3CA9-11D9-934E-003065ABFD92@mac.com> In-Reply-To: <20041121191737.GA2270@gicco.homeip.net> References: <20041121093347.GA861@gicco.homeip.net> <41A0B955.8090700@mac.com> <20041121191737.GA2270@gicco.homeip.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Nov 21, 2004, at 2:17 PM, Hanspeter Roth wrote: > On Nov 21 at 10:50, Chuck Swiger spoke: >> Hanspeter Roth wrote: >>> I have set an idle timeout for the hard-disk. But when there is no >>> user activity there are frequent disk accesses. >> >> Yes, this is Unix. Even when there is no user activity, a Unix system >> normally is still running a number of daemons such as syslogd which >> regularly write to the filesystem. Beyond that, the syncer mechanism >> tries >> to reduce the number of dirty memory buffers every thirty seconds or >> so. > > I guess that some daemons are causing disk access. But it must be > not only syslogd. That's right. Normally, people end up running a number of daemons like sendmail or some other MTA, ntpd, named, etc. > Is the syncer causing the disk to spin up even if there is nothing to > flush? Probably no. However, if you have active processes running on the system, it is very likely that the syncer will find data that it does want to write. > [...] >> Instead you probably will need to mount filesystems read-only and >> create >> RAM disks in a fashion similar to booting off limited-write media like > > My idea is to transfer those files that are written also when the > user is idle to a RAM disk (some from /var/log and dhclient.leases). > But I don't want to mount the filesystems read-only. OK. However, you are probably not going to be able to prevent everything running on a normal Unix system that wants to scribble to disk short of heroic measures. >> Compact Flash. Either that, or simply shutdown the system or run zzz >> to >> suspend the system via APM/APCI. > > This is less convenient and probably doesn't work on my laptop. (I > have to check whether the upgrade to 5.3R has changed something in > this respect.) Hmm. For what it is worth, it's taken about two years of effort by Apple to work through many of these issues in order to get MacOS X on their laptops to be reasonably friendly in terms of saving power, conserving hard drive access, and having power save/suspend to RAM behave properly. -- -Chuck
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?BE4BE0F1-3CA9-11D9-934E-003065ABFD92>