ATA port freezes randomly

Bug #157650 reported by gwi
2
Affects Status Importance Assigned to Milestone
Ubuntu
Invalid
Undecided
Unassigned
initramfs-tools (Ubuntu)
Invalid
Undecided
Unassigned

Bug Description

My system has random freezes, sometimes during boot, sometimes after a few minutes, sometimes after a few hours. Everytime there are ata port errors like "ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x1c00000 action 0x2 frozen" (see logs). I had this with Feisty, but even after a fresh Gutsy install the problem doesn't disappear.
I tried three different motherboards, two different harddisks, on the current motherboard two different SATA-controllers.

When the system freezes, I sometimes can continue after a minute or so. Sometimes I have to reboot using either the reset button or the alt-sysreq-SUB sequence. Sometimes after a reboot Grub is loading, and the gives a "Read error".

Usually after a fresh install the system runs fine for about two weeks (only running a few hours a day, then shut down). The frequency of freezes then increases up to a point where it becomes unuseable.

Revision history for this message
gwi (george-willegers) wrote :
Revision history for this message
gwi (george-willegers) wrote :
Revision history for this message
gwi (george-willegers) wrote :
Revision history for this message
gwi (george-willegers) wrote :
Revision history for this message
gwi (george-willegers) wrote :
Revision history for this message
gwi (george-willegers) wrote :

Tonight had to reboot 5 times before I could login. The other 4 times the system froze before that point. I would classify this bug as critical (if I could). The past 6 months I spent more time recovering from these freezes, reinstalling Ubuntu, than I spent doing real work (and that's what I have my computer for)!

Revision history for this message
gwi (george-willegers) wrote :

Helloooooohhhooooo, Ubuntu team, anybody there?
Again several reboots needed to get the system up and running. And even then several errors, as the attachment shows.

When is someone going to take a look at this, and decide whether or not to do anything about it?
This is a critical bug! Solve it!

Revision history for this message
gwi (george-willegers) wrote :

Can anyone tell me why this bug has been linked to initramfs-tools?

Revision history for this message
Rui Bernardo (epimeteo) wrote :

Hi had this problem with Gutsy in a qemu VM. I had to add the module ide-generic on boot:

Edit the file /etc/initramfs-tools/modules

   sudo nano /etc/initramfs-tools/modules

and add "ide-generic" in it, then run

   sudo update-initramfs -u

and reboot. No more problems since then.

Revision history for this message
gwi (george-willegers) wrote :

I am having the problem with a SATA-disk (the only harddisk in my system). Does ide-generic handle SATA as well as IDE?

Revision history for this message
gwi (george-willegers) wrote :

Some response or acknowledgment from offical Ubuntu side would be nice.

Revision history for this message
ziphyre (ziphyre) wrote :

Same problem here, and it is very strange that there is still no response or acknowledgment from anyone!!
And worse, there are plenty of people who have the same issue on ubuntuforums.
Did you happen to solve it?

Revision history for this message
gwi (george-willegers) wrote :

No, even worse. The frequency of the errors was so high that it was again impossible to do work on the computer. I decided to try Hardy Heron Alpha 1, to see if a newer kernel would solve the problem. (I read somewhere that 2.6.23 might do that.)
I tried the upgrade twice, and twice the ata errors made the upgrade freeze, and leave the system in a more or less unusable state. First time I could recover it, using a backup. Second time even that would not work.
So I tried a fresh install of Hardy Heron. But the installer used dialogs that were larger than the display, so the next/back/cancel/finish buttons were invisible. So I just finished a fresh install of Gutsy... My guess is the ata errors will slowly come back after a few weeks (as they did since about april, shortly after the upgrade to Feisty).

Revision history for this message
gwi (george-willegers) wrote :

[quote]My guess is the ata errors will slowly come back after a few weeks[/quote]
Wrong assumption: the day of the reinstall I have used the system for over 8 hours, without a problem. Used VMware server to install three more servers in the virtual network, made backup copied of the virtual systems (total more than 22GB of data). All without a problem. This morning I had to boot three times before I got to a login screen. And even in that sessions things were slowed down because of the ata errors.

Revision history for this message
ziphyre (ziphyre) wrote :

I'm downloading right now Fedora 8, and will try with it. I know it's a hopeless shot, but I have no other idea. But on the other side, I don't know how to troubleshoot and log these errors. Like you said, it's possible to go hours without a problem. The "ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x1c00000 action 0x2 frozen" error is evident, since you can't continue after. But other errors are more tricky. So if you can describe me how I can see if there are ATA errors in background, which log files should I look etc.. Then, I think I can comment more accurately on Fedora 8 behavior.

PS: By the way, my error line begins with "ata3.00: exception..." but I don't think this makes a difference?

Revision history for this message
gwi (george-willegers) wrote :

Bought a new cable last monday. It was sold as a "SATA II cable". It did not look any different than the cables I already tried, except the connector on the motherboard side has a metal clip on it, to secure the connector to the board.
To my surprise I haven't had any more ata errors since. Still can't beleave it, but if I don't see any ata error in the next two weeks, I will regard this problem solved for me...

Revision history for this message
gwi (george-willegers) wrote :

Since I replaced the SATA cable by a "SATA II" cable, I haven't seen the ata errors anymore.
So AFAIK this bug can be closed.

Revision history for this message
James Collier (james-collier412) wrote :

This bug report is being closed due to your last comment regarding this being fixed with an update. For future reference you can manage the status of your own bugs by clicking on the current status in the yellow line and then choosing a new status in the revealed drop down box. You can learn more about bug statuses at https://wiki.ubuntu.com/Bugs/Status . Thank you again for taking the time to report this bug and helping to make Ubuntu better. Feel free to submit any future bugs you may find.

Changed in initramfs-tools:
status: New → Invalid
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.