Indeed Mr. Brown is right on the money. I was able to visit the machine today, and powering down and/or rebooting without APM results in everything working fine. Shall I wonder over to linux-kernel and beg for more time in the APM power-off cycle? I take in that md gets some sort of notice from APM that the machine is going down and there simply isn't enough time in my machine. (So, md waits for APM notice, and APM should wait for some sort of status that HDs are ready for power off?) Thank you very much for the assistance. -- Brian Litzinger <brian@worldcontrol.com> > > On Thu, Mar 28, 2002 at 04:12:37PM -0800, brian@worldcontrol.com wrote: > > > You can read all the gory details of my RAID1 setup at > > > http://www.litzinger.com/raid.html > > > ... > > > I created a bootable RAID 1 setup and everything works fine. > > > Except, every time I power down the system and bring it back > > > up the system reports that the RAID array is unclean and > > > spends and hour doing a reconstruct. > On Friday March 29, brian@worldcontrol.com wrote: > > Responding to my own email, the message that flies by > > is 'invalidate: busy buffer'. There are generally > > what looks to be between 3 and 30 of them. > On Sat, Mar 30, 2002 at 09:02:08PM +1100, Neil Brown wrote: > These messages are normal if your root filesystem is on md/raid. > As the root filesystems is never unmounted, there are still buffers > active when the md devices finally gets turned off and invalidates all > of its buffers. > > It sounds like your machine is getting switched off too quickly after > the raid shuts down and your IDE drives which has said that the data > is safe, haven't actually had a chance to flush it out of cache yet. > > Can you compile without APM support so that it doesn't actually power > down itself? - To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
