Phil Endecott wrote: >>> They have used something called sReadAhead to prefetch the required >>> blocks from the disk. Hopefully this can be packaged for Debian.
> I've tried it, and it does seem to basically work in as much as I have > fewer disk access peaks in the bootchart output than I did before. But > I do still have one huge peak from fsck (is that right?) It wasn't fsck - and I recommend using bootchart -n if you see something in its output that looks suspicious. It was actually sm-notify, which is an obscure NFS thing. If you never use (or installed) NFS, stop reading now. The purpose of this thing is to notify any NFS peers from before you rebooted that you're now alive again. But in the process it calls sync() which causes lots of disk activity and slows other processes that are trying to read from the disk. And in the case where you didn't use NFS at all last session [e.g. I have NFS behind autofs to access other local machines, but use it maybe only once or twice a week], it is totally unnecessary. So I have made a small patch to sm-notify that causes it to terminate early in the case where there are no hosts to notify. You can find it here: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=502122 I get an improvement of 2-3 seconds in boot time as a result. Changing the subject: Arjan van de Van has just posted some git links to his "fastboot" kernel on LKML: http://thread.gmane.org/gmane.linux.kernel/743932 Phil. _______________________________________________ Debian-eeepc-devel mailing list [email protected] http://lists.alioth.debian.org/mailman/listinfo/debian-eeepc-devel
