2001 Quicksilver G4/867 (M8360LL/A) "locks up"

Question

Upon purchasing a 2008 Mac Pro last year, I delegated my 2001 Quicksilver G4/867 that I use on my home network to the roles of a personal mail server (5-account DoveCot IMAP and Postfix SMTP), ssh-publickey-authenticated server, limited afp file server (only a 60GB HD for five users), with limited (single user) screen sharing access, printer server, fax server, and phone logger/voice mail server. Oh, and it runs Snort network intrusion detection software, too.
So it's a busy computer. I have, however, been recently experiencing computer crashes. I personally observed the kernel crash screen once during these recent times, but usually, the screen is black by the time I am able to get to the computer to "fix" it, I presume because although I don't let the computer CPU sleep, I do allow the screen (and hard drive, when possible) to sleep. In order to recover from these computer freezes, a hard power-down is required. I follow this by a cold-boot into single-user mode. After at least two executions of /sbin/fsck -fy and a final reboot, all is back up and running again ‑‑ at least for awhile.
A review of the console message log after the most recent crash yielded the following suspect entries:
7/8/09 08 Jul 2009 13:24:27 kernel IOATAController device blocking bus.
7/8/09 08 Jul 2009 13:26:43 kernel IOATAController device blocking bus.
7/8/09 08 Jul 2009 13:26:43 kernel IOATAController device blocking bus.
7/8/09 08 Jul 2009 13:26:43 kernel IOATAController device blocking bus.
7/8/09 08 Jul 2009 13:26:43 kernel AppleATADiskQueueManager::setPowerState(0x29bf400, 2 -> 1) timed out after 100185 ms
7/8/09 08 Jul 2009 13:26:43 kernel IOATAController device blocking bus.
7/8/09 08 Jul 2009 13:26:43 kernel IOATAController device blocking bus.
7/8/09 08 Jul 2009 13:26:43 kernel IOATAController device blocking bus.
7/8/09 08 Jul 2009 13:42:30 dnschk\[761] / (O.P. note: this is a homebrew unix shell script controlled by an hourly launchd.plist -- included here only to show that the computer didn't freeze up immediately after the last 13:26:43 entry)
7/8/09 08 Jul 2009 17:13:38 kernel Darwin Kernel Version 9.7.0: Tue Mar 31 22:54:29 PDT 2009; root:xnu-1228.12.14~1/RELEASE_PPC (O.P. note: this is reboot after execution of /sbin/fsck ‑fy in single-user mode following hard power-down) 
There are no applicable crash reporter logs or panic reporter logs or hang reporter logs created during these time periods.
My questions: does this message log suggest a hard drive that needs to be replaced? Or failure of another replaceable component and if so, what might be some likely suspects? Any other suggested hints for troubleshooting such as run off my clone drive (I've been a little hesitant about cloning the main drive since I wasn't sure if some sort of file corruption on the main drive might be the culprit)?

Answer 1

The computer failed to shut down last night (it's getting up in years, so I figure it needs its beauty rest for a few hours before starting another 19-hour day).
Cold reboot following yet another hard powerdown was very difficult, and I could not boot up into single-user mode. There was another "bus blocking" nastygram and single-user mode bootup stalled immediately thereafter. Another hard powerdown and cold bootup into multi-user mode was very slow, but I eventually was able to log into an administrator account. Tried to launch Console.app and Disk Utility.app and stalled with a spinning pinwheel of death and listening to a hard drive thrash about.
I unplugged the drives one by one to see if it would boot. Trying to boot off of my backup clone drive failed. With the backup clone drive unplugged (it is the 8-yr-old OEM drive, by the way, with at least the 1st 5 yrs being the primary drive), the computer booted up fine in both single-user and multi-user modes. No "bus blocking" nastygram in single-user mode.
Several executions of /sbin/fsck ‑fy cleaned up the newer Seagate. Didn't check the console log after multi-user restart; I'll see how it behaves over the next several days. Hopefully, I'll get off easy and only need to replace that old IBM drive ‑‑ they're pretty cheap now that SATA and SSD are all the new rage.
I think my plan will be to live without "server-lite" backup for a few days and see how the old Quicksilver behaves; if it behaves over the next several days, I'll buy a new drive for bootable backup then see if it continues to play well with others. I'll post back in a couple of several days with a report on that.
Then all I'll need to do is figure out what kind of screwdriver I need in order to get those ✸☢☣⚡♨♠❦ screws out that are holding that ✸☢☣⚡♨♠❦ drive bay in place. It's not a Phillips head style of screw and it's not a Torx-head style of screw, either.
Oh, and thanks for that link to the Apple Hardware Test CD. After eight years, I have no clue where mine has gone. I'll post back with how that works out, too.