Hi, I thought I contribute some additional info to this problem.
We have about ~50 nearly identical all-in-one PCs out there that are
experiencing relatively often the exact problem described here.
The industrial-grade hardware is based on Intel Atom N270 (1.6 GHz) -
see full specs here: http://www.cartft.com/catalog/il/1325
We use different kinds of non-mechanical hard disks:
- either a Kingston 8GB SSD SATA drive (most are using these, then we tried
using other drives to avoid the problem we're talking about here)
- or a SATA-to-CompactFlash adapter with a 4GB SanDisk Extreme CompactFlash
card (we use those cards - without adapter - in other configurations and never
had any problems with them)
- or a Transcend TS8GHSD310 8GB SSD SATA drive
They are all running the same Linux image based on Debian 6, Kernel
2.6.38.6 and run 24/7. The system has been specifically modified to
avoid any unnecessary Flash writes. All partitions are r/w and
occasional writes are possible, but overall there are certainly far less
than 1 MByte writes per day.
Basically we're only using half of the 8 GB drive space so these drives
have a lot of room for wear leveling.
In the past months we had a number of failures that were mostly related
to the frozen SATA link described in this bug. The drives would not even
recover with a software reboot - the PCs need to be powered down and up
again.
We're using high quality cables and also tried to use different cables.
Note that these machines are firmly mounted, have no moving parts
whatsoever and are located in lonely places - i.e. there is no
mechanical reason why the SATA cable should suddenly have problems.
We only started recently to log in detail which hardware failed but I
can defenitely say that it happens relatively often with the Kingston
drives and it happened at least once with the CF adapter. I have no bad
records for the Transcend drives so far but only five machines are using
them so far. I wouldn't be surprised if even those fail.
My personal opinion is that there is some sort of incompatibility between the
Linux drivers and the SATA controller or SATA drives that reveals itself only
in certain rare situations. But other that that I'm clueless.
These machines most of the time do nothing but wait, so CPU time and
disk accesses are very low.
BTW, the same Linux image is used in a number of other machines without
SATA drives (they have CF card slots) and never had any comparable
problems. Also I was never able to reproduce the problem in our office,
which makes testing very hard.
I hope this helps to narrow down the problem. Let me know if I can be of any
further help.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/965213
Title:
ata1.00: exception Emask 0x50 SAct 0x1 SErr 0x480800 action 0x6 frozen
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/965213/+subscriptions
--
ubuntu-bugs mailing list
ubuntu-bugs@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
