Major startup problems with Kubuntu 10.10 Maverick

[coastalroad]

I am experiencing a serious problem that I have no familiarity with. Hoping somebody here might be able to help.

Here's the story: Earlier today, I restarted my Kubuntu machine, as I periodically do. After the manufacturer screen had come and gone as usual, I was left with a blinking cursor, which persisted for several minutes. Normally, it takes only a few seconds for the login screen to come up, but the system seemed stuck. At last something happened. Kubuntu started spitting out what appeared to me to be a bunch of kernel errors. Lines like this:

devd-work[365]: open /dev/null failed: No such file or directory

and this:

ata1.01: exception eMask 0x0 SAct 0x0 SWee 0x0 action 0x0 frozen
ata1.01 failed command: READ DMA
ata1.01: status { DRDY }
ata1.01 cmd c8/00:08:08:08:c0/00:00:00:00/50 Emask 0x4 (timeout)

... and lots of other lines beginning with numbers and then immediately followed by "ata1.01". I couldn't write them all down because they eventually disappeared.

A sequence of errors like this cascaded several times before the Kubuntu blue screen suddenly showed up, and was replaced in turn by the login screen.

I attempted to login, and KDE loaded, but then it froze. I used the magic sysreq combination to get the computer to restart. But the same thing happens. Blinking cursor for several minutes. Then several minutes worth of error messages. Then I see the login screen. Now I can't even login. Kubuntu freezes when I try to do so.

Something is seriously wrong. That much is evident. But what is wrong, I don't know.

I tried booting into the Ubuntu Live CD (using an actual CD) but it got stuck after I clicked "Try Ubuntu". After about ten minutes I used the magic sysreq keys to again force a restart.

I next tried a USB thumb drive. Miraculously, this worked. I got in and I checked my disks using the Disk Utility. The SMART test reported the SDD that Kubuntu runs on as healthy. A filesystem check (run with the drive unmounted) also produced no errors.

I restarted again, hoping the problem might have fixed itself. It hasn't. And I am unsure as to what my next steps should be. Advice?

[dibl]

Re: Major startup problems with Kubuntu 10.10 Maverick

It appears to be all about the hard disk drive. Sorry.

Can you boot a Live CD and run fsck?

With the hard drive NOT mounted, from a live CD session, and assuming it is /dev/sda1 :

sudo e2fsck -pv /dev/sda1

Good luck.

FYI, SMART can report good news on a dying drive, and bad news on a healthy drive -- it's not all that reliable, unfortunately.

[coastalroad]

Re: Major startup problems with Kubuntu 10.10 Maverick

It appears to be all about the hard disk drive. Sorry.

Can you boot a Live CD and run fsck?

With the hard drive NOT mounted, from a live CD session, and assuming it is /dev/sda1 :

sudo e2fsck -pv /dev/sda1

Good luck.

FYI, SMART can report good news on a dying drive, and bad news on a healthy drive -- it's not all that reliable, unfortunately.

Done:

ubuntu@ubuntu:~$ sudo e2fsck -pv /dev/sda1
/dev/sda1: clean, 194460/977280 files, 1906818/3906048 blocks
ubuntu@ubuntu:~$

The drive's behavior has not changed.

I actually have /home mounted to an entirely different drive - or set of drives, I should say. My RAID controller card says the array is healthy, but the controller card's alarm has gone off twice since I started experiencing these errors. I unplugged the card to see if it would make any difference. It doesn't. The same errors still show up onscreen. It seems to be a problem with the SDD that's running the system. But I don't know what's causing the Adaptec controller card to emit high-pitched noises. As far as I can tell, nothing else seems to be wrong. Any other tests I should run?

[dibl]

Re: Major startup problems with Kubuntu 10.10 Maverick

I don't know the specs for your Adaptec card, but everything you are describing indicates a hardware failure of some kind -- either the controller card or a hard disk drive. At this point, your focus needs to be 100% on data backup.

Once your data is safe, I would start taking apart the hardware, and testing each component individually.

Actually, if it were my data, I'd be replacing all of the storage components -- I can't stand the risk to my data of failing hardware.

[coastalroad]

Re: Major startup problems with Kubuntu 10.10 Maverick

I don't know the specs for your Adaptec card, but everything you are describing indicates a hardware failure of some kind -- either the controller card or a hard disk drive. At this point, your focus needs to be 100% on data backup.

Once your data is safe, I would start taking apart the hardware, and testing each component individually.

Actually, if it were my data, I'd be replacing all of the storage components -- I can't stand the risk to my data of failing hardware.

Well, the problem is with the SSD. I know there's nothing wrong with either the Adaptec card or the HDDs it manages. The alarm can be triggered by conditions other than disk failure.

The SSD, on the other hand, appears to have gone bad. Fortunately, it is still under warranty, and I will contact the manufacturer to get a refund or a replacement. Because /home is not mounted on that drive, reassembling my installation should not be too difficult.