I’m suddenly experiencing a very annoying issue for a few days (which also got me to think my system broke and won’t boot any more). After GRUB and before the splash screen shows, I’m stuck in a console printing errors about my sata drive for 5 minutes. After this booting resumes as normal. It only appears to happen if I restart my computer with the restart command or a hard reset, and not if I power it off for a few seconds then turn it back on. The message getting spammed every 15 seconds is:
ata17: COMRESET failed (errorno=-32)
ata17 was the last time, at each boot that number changes. I googled about this but nothing was precise. Some people said it means the hard disk is failing, which is NOT true. I have SMART enabled in BIOS to check that, and never had any read or write errors on my drive. My drive is also a Seagate which I know are of good quality and durable, plus I never had issues with it in nearly 3 years (2TB SATA3 drive). Other websites said it’s a BIOS problem, but what is the actual component / setting? I run my drive on a SATA3 port and everything is set to AHCI (no RAID). Does anyone know how to get rid of that message so I won’t have to wait 5 minutes after restarting?
I am not sure how to get rid of an error message about your three year old hard drive that has been working properly until up to now. My thought is that all hard drives fail, back up often and all things work great until the time and place of their own choosing when they decide to fail. I have had several hard drives that failed way before then and still under warranty. It does depend on how many you use, but the fact that you can still use it may be a blessing letting you back up everything before it goes out completely. In the mean time you can use my FastBoot bash script to restart without rebooting:
Thanks… but the hard drive isn’t failing. I’m very sure of it. I only started getting these problems in SUSE (Windows works fine) and don’t have the slightest issue with my drive otherwise. I suspect it’s the Kernel having slight issues in understanding SATA3 or something. Isn’t there some flag or option to disable whatever check causes those messages to happen? Or how can I safely test what else might be causing it? Have other causes been known?
Played around with AHCI / SATA3 modes in BIOS earlier. Nothing seems to be fixing this issue, and the hard drive works perfectly well otherwise.
It’s clearly something openSUSE is trying to do / check about my drive and not succeeding, which I likely don’t need it to do. At the end it fails all attempts and just moves on as usual. Is there any safe way to disable whatever openSUSE does that causes this? It’s really annoying to deal with something like this at each startup.
Something interesting I noticed (but am not sure about): If I startup my computer after it’s been powered off, or restart it in openSUSE, this issue happens. But if I boot into Windows 7 first and use the restart command there (without even logging on), next time it boots openSUSE no longer shows those errors. The problem doesn’t happen all the time either, just most of the time. Maybe it’s related to how the system gets restarted or some weird stuff, though I don’t see how a restart can be different based on what’s triggering it.
MirceaKitsune wrote:
> I’m suddenly experiencing a very annoying issue for a few days (which
> also got me to think my system broke and won’t boot any more). After
> GRUB and before the splash screen shows, I’m stuck in a console printing
> errors about my sata drive for 5 minutes. After this booting resumes as
> normal. The error only appears to happen if I restart my computer with
> the restart command or a hard reset, and not if I power it off for a few
> seconds then turn it back on. The message getting spammed every 15
> seconds is:
>
>
> Code:
> --------------------
> ata17: COMRESET failed (errorno=-32)
> --------------------
Is that all you’re seeing? That isn’t the full error message and in
particular doesn’t include details to help diagnose the problem. If that
is all you’re seeing, please search for the full message. Perhaps
‘dmesg’ will find it or it may be in a system log.
Every 4 lines of this type it shows a few more things, but it’s not an interactive console so I can’t print it to a file or anything. Do these logs get stored in a text file? If not I’ll take a picture of the screen with my mobile phone and post that.
Thanks. Might try that script tomorrow. After looking more at this issue, I’m getting the impression it’s actually not the hard drive, but an USB device (no idea which, I have a lot connected). I was confused by the ata part and confusing that with sata, but pretty sure it’s not that. My senses tempt me to think it’s the webcam, which also doesn’t work in openSUSE.
MirceaKitsune wrote:
> Thanks. Might try that script tomorrow. After looking more at this
> issue, I’m getting the impression it’s actually not the hard drive, but
> an USB device (no idea which, I have a lot connected). I was confused by
> the ata part and confusing that with sata, but pretty sure it’s not
> that. My senses tempt me to think it’s the webcam, which also doesn’t
> work in openSUSE.
No it’s an ATA device. That’s because the error message says ata and not
usb!
Funny thing: I booted Clonezilla LIVE yesterday to make a partition backup, and after it started up it went on spamming the same message all over the interface (just like openSUSE when booting). Had to boot into Windows, click restart there and it stopped. This is silly… as again I checked everything and neither the BIOS settings nor the hard drive should have any problem. All I can do is restart as little as possible and hope it eventually stops.
A worth revive of this thread after nearly 4 months. I decided to attend the problem again today, since I got the impression it was contributing to slowdowns and freezes in the OS as well. At last I found a way to fix it… still no idea what exactly was causing it though.
The issue was likely not a hard drive, but the SATA chip on the motherboard. My mobo comes with two: eSATA and GSATA. By default both are enabled but either can be turned off from BIOS. I disabled the eSata controller entirely, and that’s when the issue stopped. Both of my hard disks and DVD-RW are still being detected, I’m still running them in AHCI mode without any problem, and those messages and boot slowdowns are completely gone from startup.
My conclusion is that there’s a major problem with the eSATA chip on my motherboard. Since the mobo came in perfect condition I don’t think it’s hardware related, but a badly programmed firmware. By the time I got this mobo SATA3 (not sure if related to eSATA) was still experimental and this was a very early implementation. But yeah… if anyone else ever experiences this, check if you can toggle / change SATA chips in BIOS, and if you use eSATA disable it.
MirceaKitsune wrote:
> My conclusion is that there’s a major problem with the eSATA chip on my
> motherboard. Since the mobo came in perfect condition I don’t think it’s
> hardware related, but a badly programmed firmware. By the time I got
> this mobo SATA3 (not sure if related to eSATA) was still experimental
> and this was a very early implementation. But yeah… if anyone else
> ever experiences this, check if you can toggle / change SATA chips in
> BIOS, and if you use eSATA disable it.
What model of chip is it? For that matter, what is the other chip as
well. If it is a firmware problem then perhaps there’s a BIOS upgrade.
But hardware design errors are not unknown. More recent versions of the
driver may have workarounds.
I don’t know that. But the specs for the motherboard should be out there… it is a GA-X58A-UD7 rev 1.0. Here is a link to its official page in case it’s listed there: GIGABYTE - Motherboard - Socket 1366 - GA-X58A-UD7 (rev. 1.0) As for the bios, the last stable one is F7 which I already have.
All I know is that there are two sata controllers, eSATA and GSATA. Both are normally enabled in BIOS. But if I turn eSATA off and leave only GSATA, the problem no longer happens.
Also: I believe it might not be a Linux specific problem. While this issue was taking place, the computer would spend a longer amount of time during the boot process for the “Detecting hard drives” step (before reaching GRUB), which felt like there’s an issue. It probably means the reason it didn’t happen in Windows was that Windows handled it in such a way it wasn’t noticeable.
Another thing that remains a mystery is why my PC would startup with the drive / chip in a buggy stage if I powered it off or restarted from Linux, but not if I issued the restart command from Windows (in which case it would no longer happen during that startup). Since no sort of information is stored anywhere in either case. Restarting from Windows 7 without powering off was the only thing that would allow me to boot into openSUSE without this problem (I removed Windows so my machine is no longer dual-boot).
MirceaKitsune wrote:
> djh-novell;2516813 Wrote:
>> What model of chip is it? For that matter, what is the other chip as
>> well. If it is a firmware problem then perhaps there’s a BIOS upgrade.
>> But hardware design errors are not unknown. More recent versions of the
>> driver may have workarounds.
>
> I don’t know that. But the specs for the motherboard should be out
> there…
Well, if you can’t be bothered reading the spec, neither can I. I
haven’t bothered reading the rest of your message, either.
Um, I don’t even know how to get the firmware and chip versions for these (which isn’t a general front-page spec of any motherboard IIRC). Maybe I will look more thoroughly for it on google later, if I understand what exactly to search for. Gotta love how everyone assumes that if you use Linux, you know or can find any deep technical data in a matter of seconds, even when you’re completely new to something
On 01/10/2013 04:36 AM, MirceaKitsune wrote:
>
> djh-novell;2516946 Wrote:
>>
>> Well, if you can’t be bothered reading the spec, neither can I. I
>> haven’t bothered reading the rest of your message, either.
>
> Um, I don’t even know how to get the firmware and chip versions for
> these (which isn’t a general front-page spec of any motherboard IIRC).
> Maybe I will look more thoroughly for it on google later, if I
> understand what exactly to search for. Gotta love how everyone assumes
> that if you use Linux, you know or can find any deep technical data in a
> matter of seconds, even when you’re completely new to something
Do you know what the eSATA chip/port is used for? I doubt that you do as you
expressed surprise that everything still worked after you disabled eSATA.
The meaning is “external SATA”. Your MB has two ports on the back that are
combination USB/eSATA, where you should be able to either plug in an external
SATA disk or a USB device. If you have no such eSATA device (likely), then
disabling it was the correct thing to do.
At one point I had an MB with an eSATA port, and an external drive that used
that port. The whole setup was unstable, and I soon pulled the driver out of
that enclosure and mounted it internally. The enclosure also had a USB port, and
that is the only way I use it now.
Ah… now I remember about that yes. I don’t use any drives on the external SATA ports which is why I noticed no difference with eSATA disabled. Still, I would rather find what’s causing the problem and fix it properly than disabling the external SATA ports permanently… though I fear this is an issue too advanced for me. Sadly I have no more info on this currently, nor know how to discover the info that djh-novell was upset at me for not knowing.
