On Mon, Oct 14, 2002 at 12:54:55PM -0400, christophe barbe wrote: > On Mon, Oct 14, 2002 at 05:22:45PM +0100, Nyk Tarr wrote: > > > In my opinion this is not a solution. First I am still not convinced > > > that it is normal for ext3 to touch the disk when there is no other > > > activity on the system. Then spinning the disk for 5 minutes seems very > > > bad from the powersaving point of view and also from the life of the HD. > > > > Hard drives sustain most of their wear and use most power on spinning up, > > it makes sense for most systems (ie desktops/servers etc) to keep the disks > > spinning all the time. In the majority of cases, then, it isn't a case > > of 'spinning up' the disks at all. This is not the case on laptops, but > > we are a minority case. Optimising the commit intervals has been > > suggested on lkml, but is, apparently rather tricky. > > The commit of what? I don't see why there would be a need to commit > something on an idle system. A commit every 5 secondes when the system > (and the HD) is used is not a problem. >
It seems nothing is committed unless there is a dirty block ie something, somewhere has changed, however small or seemingly insignificant. It would be pretty difficult to stop _all_ disk activity, unless a very high percentage of your processes are running elsewhere - something like a ssh to another box with few processes still running locally. You'd have to stop *logd for example. See the discussions at: http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=linux.kernel.20011022124751.C5146%40turbolinux.com&rnum=6&prev=/groups%3Fhl%3Den%26lr%3D%26ie%3DISO-8859-1%26q%3Dlaptop%2Bmode%2Bcommit%2Bgroup%253Amlist.linux.kernel and http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&threadm=linux.kernel.3CFD453A.B6A43522%40zip.com.au&rnum=1&prev=/groups%3Fhl%3Den%26lr%3D%26ie%3DISO-8859-1%26q%3Dlaptop%2Bmode%2Bgroup%253Amlist.linux.kernel -- /__ \_|\/ /\
