
| This forum is proudly powered by Scientific Linux 6 | SL website Download SL Help Search Members |
| Welcome Guest ( Log In | Register ) | Resend Validation Email |
![]() ![]() ![]() |
| jgennaro |
Posted: Oct 11 2011, 06:54 PM
|
|
|
SLF Newbie Group: Members Posts: 7 Member No.: 415 Joined: 11-July 11 |
Hello,
Two workstations running SL6 have been operating fine for the last 3 months. My group communicates with them over SSH, but sometime yesterday evening both machines became incommunicable. Upon investigation, both machines can POST and load the SL6 os, and can successfully boot into run level 3 via GRUB. However, we are unable to get GNOME running; e.g. startx from the prompt in run level 3 results in a black screen and unresponsiveness. I have had issues with the X configuration before, which were resolved by using the ATI utility aticonfig; when things weren't working the symptoms were very similar to what we are seeing now. However I don't understand why would both machines would stop working in the same way at around the same time when they've been working just fine non-stop for the last couple months. They are both housed in the same facility which claims to have had a power interruption last weekend; is it possible that some hardware was damaged (e.g. the video cards)? What steps can I take to diagnose this? I'm looking for some help about where to get next, starting from the prompt at run level 3 which seems to work fine. Thanks, Joe |
|
| redman |
Posted: Oct 12 2011, 10:32 AM
|
|
![]() SLF Admin ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() Group: Admins Posts: 1669 Member No.: 2 Joined: 8-April 11 |
SL6 has by default yum-autoupdate enabled.
That means that any update is automatically downloaded and installed. It is possible that an update caused this problem. -------------------- What is SL? - Forum Rules - Info on 3rd Party Repos - How to post images - How to post large text / config files
Desktop: Asus P5QPL-AM, Intel Dual-Core E6500, 4GB DDR2, Asus GeForce GT 430 1GB, SL6.4 x86_64 Test box: Intel S5000PSL, 2x Intel Xeon E5310, 8GB ECC DDR2 FB-Dimm, Asus GeForce GT 220 1GB, SL6.4 x86_64 |
|
| adebened |
Posted: Oct 13 2011, 06:17 PM
|
|||
|
SLF Member ![]() ![]() ![]() Group: Members Posts: 93 Member No.: 21 Joined: 11-April 11 |
Hello, Do you have the drivers from ATI installed? I'm wondering if maybe your ATI graphics card drivers need to be recompiled after kernel updates, like the ones from Nvidia. As Redman suggested, it's possible that you had an automatic update, and that maybe there was a kernel update in there. With the power outage your computers were rebooted with the new kernel but X wont start due to not having the drivers compiled for the new kernel. In the off-chance that this is the issue, installing ati kmod drivers from elrepo would prevent this from happening again. I'm just guessing that this might be the issue though. (Also see this thread)). EDIT:
Sorry, I missed this point on first reading. It's really unlikely that what I mentioned is the issue then, as ssh would still be available if the graphics drivers didn't load. |
|||
| jgennaro |
Posted: Oct 16 2011, 10:55 PM
|
|
|
SLF Newbie Group: Members Posts: 7 Member No.: 415 Joined: 11-July 11 |
Ok, thanks for your responses.
To fix the issue we're going to make a special trip to this location; we have basically one day to ensure these workstations are going to work during the extent of our imminent scientific experiment (which is a balloon flight). Beyond the thread linked by adebened (thanks!), are there any other steps needed in SL6 to absolutely ensure there are no autoupdates whatsoever of any kind? If this occurs during our flight, we will endanger our instrument. We chose SL6 for stability so please help me ensure this is the case. About the video drivers - yes the ATI proprietary drivers are installed. However moving the xorg.conf configuration file on one of the workstations (and not replacing it with anything) allows it to start X successfully. Is there a way to 'play it safe' and just use some default drivers that will have zero chance of causing any kind of issue? Is this one way to do that? (The other workstation evidently has some kind of hardware damage - it now cannot even POST). Thanks, Joe |
|
| adebened |
Posted: Oct 17 2011, 11:46 PM
|
|||
|
SLF Member ![]() ![]() ![]() Group: Members Posts: 93 Member No.: 21 Joined: 11-April 11 |
Hello, Sorry for my delay in replying. I use NVidia so I don't have much experience with ATI but I am guessing (hopefully that someone who uses the ATI drivers and knows for sure can verify) that the elrepo repositories have an ATI driver kmod. The advantage of the kmod is that, if you update the kernel, you do not need to recompile the drivers when you boot into the new kernel (which, with the driver binary provided by NVidia, you do). They are the same drivers, however. I can't find any packages with "ati" in their name in elrepo but maybe they are going under a different name? If you now have the autoupdate turned off there should be no more surprises... hopefully |
|||
![]() |
![]() ![]() ![]() |