It's one of those things. Yes, of course you are right. I have one on my servers, and it's great until it beeps at 2am! However, until I did my annual testing on the batteries, I didn't realise they were completely dead! (There's someone on trademe who does replacements, and he's extremely helpful. I'll dig out my receipts if someone's after a contact).
This site does have a UPS - a small one, but it couldn't provide protection against this spike. As most can handle about 50,000V, that's a bit scary! We could do with someone (Volker?) to explain the physics of the problem... Steve. On Thu, 27 Mar 2008 07:58:45 +1300 David Lowe <[EMAIL PROTECTED]> wrote: > > Unless the cement wagon takes out another power pole, of course. > > At risk of stating the obvious I'm wondering if a small investment in little > UPS might be in order. I live rural, and I can attest that power supplies > anywhere outside the big smoke are flakey at the best of times. > > An interesting thread thanks. > > - D > > -----Original Message----- > From: Steve Holdoway [mailto:[EMAIL PROTECTED] > Sent: Wednesday, 26 March 2008 9:15 p.m. > To: [email protected] > Subject: Re: Any guru near Governers Bay? > > On Wed, 26 Mar 2008 19:51:43 +1300 > Volker Kuhlmann <[EMAIL PROTECTED]> wrote: > > > On Wed 26 Mar 2008 17:42:52 NZDT +1300, Steve Holdoway wrote: > > > > > That's what I'm going to do. As not that many distros use lilo any > > > more, I'll be chrooting to the old system first. > > > > My guess is that someone updated the kernel and forgot to reinstall > > lilo. That's more than a good enough reason to ditch lilo and use grub, > > the current problem wouldn't have existed in the first place. > > > > Volker > > > Couldn't agree more! Still only partially fixed... but it's up and running, > which is the important part, as it runs important systems, mail etc... > > Turned up with 3 disks to boot off, FC7 live CD, Knoppix 5.1 DVD and an > AMD-64 ubuntu one just in case. Server only had a cd drive, so Knoppix was > out. Booted up no problem on FC7 live, and found 2 disks which would > normally have been a softraid pair, but weren't. So fsck'd both disks, and > mounted the first root partition. This is when the fun started... > > mkdir /mnt/a > mount /dev/sda1 /mnt/a > chroot /mnt/a /bin/bash > seg fault > > B*ll*cks! Spent about 45 minutes going nowhere trying to get chroot to work > with various shells, then trying to run the mandrake lilo (as FC7 doesn't > use it any more), fighting with LD_LIBRARY_PATH and the like. But it was a > waste of time. > > I had noticed, but not registered that the 2 hard disks were in removeable > SATA caddies. Aha! let's try splitting the mirror to see if one of the disks > has an uncorrupted bootstrap, just get them back up and running! No key ): > > In the end it was a bonnet up job, and disconnect a disk from the > motherboard ( the right one first time! ) and get it booting. As I use the > same caddies myself, I'll take a key over tomorrow and put it all back > together. I'll leave it there, too (: > > As you said Volker, a properly set up grub would have a recovery boot option > off either disk as well as the mirrored pair ( and I would have got the menu > ), and I sure the lady in charge could have been talked through the process > of an emergency boot. The risk of running unmirrored for a short period of > time was certainly acceptable, even to the eternal pessimist - myself. > > Unless the cement wagon takes out another power pole, of course. > > Steve > -- > Steve Holdoway <[EMAIL PROTECTED]> >
pgpxe8q5IIdQY.pgp
Description: PGP signature
