DebianThis forum is for the discussion of Debian Linux.
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
I did a apt-get dist-upgrade today and immediatly noticed that apache, squid, php4 and mysql stuff wasn't working. I've fixed most of those problems and rebooted the server just to make sure everything would come back up right.
The server simply boots to a screen saying LI.
I figured i probably need to get in with a rescue cd and check the lilo config and run lilo -v, so thats what I've been trying.
I can't get any rescue cd to boot for me.
I've tried a debian 2.2 cd, debian 2.4 cd and knoppix so far.
I have a Mylex AccelleRaid 160 (DAC960) raid controll, and I think thats whats holding me up.
with both debian CD's I've been trying the following:
First the compliments: Too few posters have the courtesy or thoughtfulness to add additional info as you did. Hope this softens the criticisms.
It helps to spell correctly, especially the name of the hardware involved:
"Mylex Acceleraid 160"
Fortunately, Google suggested the above correction.
For those of us who are not familiar w/ this item, a fuller description, perhaps "Mylex AcceleRAID 160 PCI Ultra-160 SCSI RAID LVD" (right out of the Google results), would be helpful. I guessed the "PCI" & "SCSI" parts, but ...
What else have you tried?
Your original post details the problem & your attempted solutions well. Have you done any Google, LQ, or other searches to find a solution?
Suggestions / Questions
Is a re-install feasible?
What was setting up the card in the original install like? Any insight there as to what might be wrong now?
And thats it....the cursor blinks at me and just sits there.
And I assume is unresponsive.
Rumor, collected wisdom, & one horrible personal experience say that
can hose a system. I will never do one again until: it is easily reversible, there are better safety checks, or I understand the process way, way better.
Rumor, collected wisdom, & one horrible personal experience say that
can hose a system. I will never do one again until: it is easily reversible, there are better safety checks, or I understand the process way, way better."
I can't remember the last time i had a problem with 'apt-get dist-upgrade'. Of course, i'm one of those nuts who read the APT-HOWTO until practically it was second nature.
"I did a apt-get dist-upgrade today and immediatly noticed that apache, squid, php4 and mysql stuff wasn't working. I've fixed most of those problems and rebooted the server just to make sure everything would come back up right.
The server simply boots to a screen saying LI."
Did you do an 'apt-get update' before the dist-upgrade?
Was Lilo updated? Did you run #lilo?
Lilo is giving you an error "LI", what does google.com/linux tells you about it?
I had done apt-get update immediately before the dist-upgrade. There were many packaged "held back" and figured the dist-upgrade would take care of that. I didn't look to see if lilo was updated and didn't run lilo before rebooting.
I've tried everything I can find via google and several forums, but havn't had luck yet. I've found instances of this exact same problem on completely different configurations, but every forum I've seen the problem posted did not have any responses regarding a solution.
I've been running linux for a few years, but tend to just keep the box updated and am still learning from my mistakes or jumping into something I should have read about or at least paid closer attention too.
Re-installation is an option, but I havn't backed up my home directory in ages and would really like to get that data back at the very least.
The first time I installed on this system I had no problems, my debian 2.4 CD's kernel is configured to support the raid controller. It was oddly enough the smoothest installation I've done so far. I had the box up and running like I wanted within a couple of hours. My previous attempts took 2 days and alot of frustration.
I'm considering removing the raid controller and drives for the time being, install onto an IDE drive and configure the kernel for the raid controller. I'm assuming once I get an OS working I can reinstall my raid controller and drives, boot from the IDE drive and get to the raid to backup my data and hopefully fix the issues with lilo and whatever else might be going wrong.
I have had that problem with lilo as well. Think I reinstalled or used the rescue CD to fix the lilo.conf file. I'm not sure how to do that on a raid hds or so.
Thing is when packages are being kept back you should be careful with them. If I do upgrade I have packages being kept back. If I then do dist-upgrade they're not kept back but I now get a list of packages that will be removed.
Also if you upgrade lilo always keep the old config file and make sure it's run without errors. That is check if it's still ok afterwards and then run "lilo".
Since my last post I've upgraded the bios, firmware, ezassist software, bootblock and diag portions (all flashable items) on my raid controller and I still get the same problems. Knoppix will boot just fine if the raid controller is not in the server.
I'm left with the question of why is knoppix remotely affected by the presence of my raid controller/debian install. Its a live cd that should work independant of other software on a system. At the very least it should begin decompressing linux and boot partially before even looking for hardware on the system.
Does anybody know whats happening before a system begins decompressing linux? It seems apparent that the system is basically locking up because the booting OS is doing something to confuse some part of the system.
I installed another SCSI HD on the onboard SCSI controller and left the Mylex Raid controller installed. I can again boot the installation of debian. From the installer I can mount the partions on the raid array to /target. But I can't run lilo. It tells me something along the lines of the boot partiton must be mounted in /target. It was mounted there, but lilo still won't work.
If I try booting with a rescue option and setting root=/dev/c0d0 or /dev/c0d0p1 it begins to boot and says kernel panic: could not mount root fs.
It does load the drives for the raid controller, see the logical array and the partitons within it. Now it just won't mount and I don't understand why. The drives are in ext3, even if the kernel isn't compiled for ext3, ext2 should still be able to read and write to the partion right?
I've read several howto rescue type documents, but they are all geared towards IDE and I get confused on which partions on my configuration I should be doing what with. Its been 2 years since I setup this box and I don't have a clue which of the 4 partions are spose to be mounted where. I just know I have a partion for /, /home, swap and something else...possible /boot, but I'm not sure.
I managed to get the disks mounted in the correct order and lilo -v would not work, but in the debian installer I hit 'Make system bootable' and it reinstalled lilo. Removed the blank HD I had installed just to make it boot and started it back up with no problems. There are still a few issues with Apache and what not, but the system is finally booting and I can sit at my desk and SSH to fix the remaining problems.