Wild Dog / Jaunty regular crashing

[corey.abshire]

Yep, I had noticed that too. I'll have to go fork over some money for a network cable long enough to stretch across my living room, but yes, I'll try it if you think it might help.

Ok. I blacklisted ath5k and also physically removed the card (logs attached showing no ath5k in prior 2 restarts). Will report back whether it has an impact soon. Thanks. - corey

[corey.abshire]

Another crash, with ath5k blacklisted and wifi card physically removed. Logs attached.

Also, I had sar running as mentioned earlier. I didn't see anything out of the ordinary in it so I turned it back off since it was causing a bunch of noise to be generated in the logs.

So, not sure where to look next.

This sucks. This machine has crashed around once a day ever since I've had it. I've had several linux boxes before this and none of them did this. I can see several other threads related to frequent random crashes in the System76 forum - is this a common thing with System76 machines? Does anyone have any data as to how prevalent this issue is, particularly with Wild Dogs?

Thanks. - corey

[corey.abshire]

Is it worth investigating whether this is some sort of watchdog failure?

corey@system76-pc:~/projects/nocrash/sar$ ps -ef | grep watchdog
root 5 2 0 21:03 ? 00:00:00 [watchdog/0]
root 8 2 0 21:03 ? 00:00:00 [watchdog/1]
root 11 2 0 21:03 ? 00:00:00 [watchdog/2]
root 14 2 0 21:03 ? 00:00:00 [watchdog/3]

I can see 4 watchdogs running right now. I'm assuming there's one process per CPU, but can anyone confirm that's how it works?

Since I was using my machine the last time it crashed, and it didn't appear to have anything wrong with it, it just restarted for no reason, one theory here could be that for some reason the process that is supposed to be writing to /dev/watchdog all the sudden decides to not do that any more, and so the kernel decides its time to restart.

Maybe something is causing watchdog to stop writing, or maybe the process itself just fails. Not sure, need some help on that.

Also, I tried to find /etc/watchdog.conf to see how its configured, and tried to check the man for watchdog from within my terminal but no luck on either front. Anyone have any idea how its configured on the Wild Dog? Also, I read something about hardware watchdogs - anyone know if the Wild Dog has one?

Thanks. - corey

[corey.abshire]

Found this in the logs:

[ 24.681259] iTCO_vendor_support: vendor-support=0
[ 24.682225] iTCO_wdt: Intel TCO WatchDog Timer Driver v1.05
[ 24.682306] iTCO_wdt: Found a ICH10R TCO device (Version=2, TCOBASE=0x0460)
[ 24.682349] iTCO_wdt: initialized. heartbeat=30 sec (nowayout=0)

So apparently it is a hardware watchdog.

Still searching for clues on how exactly it works and whether it could be an issue.

[thomasaaron]

Cory,

We have this problem on about five Wild Dogs (out of who knows how many hundreds out there). The problem is, so far I've not been able to discover a common link. All of the logs are substantially different.

So, it doesn't *appear* to be a *common* problem.

Also, it doesn't appear that you have just one thing going on, which further complicates figuring it out. For example, in one of your previous logs, you had this...

GLib-GObject-WARNING: IA__g_object_get_valist: value location for `gchararray' passed as NULL

...just about the time of the crash. I'm not sure if that's the issue, but I've definitely seen in in connection with freezes before.

Unfortunately, it's not in your most recent log.

Would you like for us to bring your system in and have a look? I'm happy to do so. The only other thing that comes to mind is re-imaging. But who wants that hassle?

[corey.abshire]

Cory,

We have this problem on about five Wild Dogs (out of who knows how many hundreds out there). The problem is, so far I've not been able to discover a common link. All of the logs are substantially different.

So, it doesn't *appear* to be a *common* problem.

Also, it doesn't appear that you have just one thing going on, which further complicates figuring it out. For example, in one of your previous logs, you had this...

GLib-GObject-WARNING: IA__g_object_get_valist: value location for `gchararray' passed as NULL

...just about the time of the crash. I'm not sure if that's the issue, but I've definitely seen in in connection with freezes before.

Unfortunately, it's not in your most recent log.

Would you like for us to bring your system in and have a look? I'm happy to do so. The only other thing that comes to mind is re-imaging. But who wants that hassle?

I'd certainly be willing to consider letting you have a look if you think that will help. I would even be willing to consider reimaging if I had a firmer understanding of what that could help (seems like the software would just get itself back into the same state its in right now).

I need to leave right now to go out of town for a few days, and will give it some thought while I'm away. I'll let you know when I get back. If you or anyone else there has any other things for me to try before shipping it back or reimaging please go ahead and post it in the meantime. Also, thanks again for your help. - corey

[corey.abshire]

I'd certainly be willing to consider letting you have a look if you think that will help. I would even be willing to consider reimaging if I had a firmer understanding of what that could help (seems like the software would just get itself back into the same state its in right now).

I need to leave right now to go out of town for a few days, and will give it some thought while I'm away. I'll let you know when I get back. If you or anyone else there has any other things for me to try before shipping it back or reimaging please go ahead and post it in the meantime. Also, thanks again for your help. - corey

Hi Tom - I'm back now, and I have thought of a few questions and concerns about your offer:

1. Cost

Would you anticipate there being any cost to me as part of the fix, or would this be covered under the warranty. Also, who pays for shipping?

2. Shipping

As mentioned above, who pays? Also, due to the other shipping issues we've seen on the forum, I would definitely be wanting insurance both ways. Who pays for that?

Also, I have already discarded my original packing material. Would system76 be sending empty material for me to repack, or should I take it for instance to a UPS store for packing.

3. Timing

How long do you anticipate having my computer after it arrives? My fear hear is sending it off and having it gone for months with minimal communication.

4. Loaner

Is system76 prepared to send a replacement, equivalent, Wild Dog for me to use in the interim? This would make it easier to cope with not having it hear for a while. Also, we could reuse the shipping materials back and forth.

5. Methodology

What else do you plan on testing that I can't test here? For instance, are hardware issues suspected to any degree. If so, would you be making any such replacements as part of the warranty, and would you be in communication with me before doing so.

-

Thanks again for the offer. I look forward to you're response.

- corey

[thomasaaron]

Hi, Corey.

Have you gotten any freezes using the usb keyboard I sent and a usb mouse (i.e. no wireless?)

Have you checked to see if you get freezes without using gnome-do?

I would like to have both of those verified before bringing the computer in.

1. Cost

Would you anticipate there being any cost to me as part of the fix, or would this be covered under the warranty. Also, who pays for shipping?

No, you're still under warranty. Unless there is user damage.

2. Shipping

As mentioned above, who pays? Also, due to the other shipping issues we've seen on the forum, I would definitely be wanting insurance both ways. Who pays for that?

Also, I have already discarded my original packing material. Would system76 be sending empty material for me to repack, or should I take it for instance to a UPS store for packing.

We would pay for shipping, and we always insure packages. However, insurance actually paying up is usually dependent on proper packing. We do not provide packing materials, but we will provide detailed packing instructions.

3. Timing

How long do you anticipate having my computer after it arrives? My fear hear is sending it off and having it gone for months with minimal communication.

It would probably be more like a week or two.

4. Loaner

Is system76 prepared to send a replacement, equivalent, Wild Dog for me to use in the interim? This would make it easier to cope with not having it hear for a while. Also, we could reuse the shipping materials back and forth.

We don't provide loaners.

5. Methodology

What else do you plan on testing that I can't test here? For instance, are hardware issues suspected to any degree. If so, would you be making any such replacements as part of the warranty, and would you be in communication with me before doing so.

First we have to reproduce the problem. In your case it should be pretty easy based on how often it has occured during the course of this thread.

Most likely re-imaging the machine and installing the drivers from the ppa would be the first line of attack. If that doesn't fix the problem, it most likely is hardware.

Then, we'd start swapping things out: power supply, graphics card, cpu, motherboard, etc...

Any hardware replacements would be covered under warranty, unless they were deemed to be user-inflicted damage.

[corey.abshire]

I had a crash just now. Logs attached.

Have you gotten any freezes using the usb keyboard I sent and a usb mouse (i.e. no wireless?)

I wasn't aware you sent a USB (non-wireless) keyboard. I have not received it yet if you have. I don't have one to try unless I go buy one or get the one you've sent. Can you please send a tracking number via email?

I already have a USB (non-wireless) mouse I can try. I can go ahead and start using that instead if you want me to test that before I get the keyboard.

Have you checked to see if you get freezes without using gnome-do?

I don't have that program installed at all (just confirmed) and wasn't even aware of it until you mentioned it here.

I would like to have both of those verified before bringing the computer in.

Will do, as soon as I get the keyboard.

First we have to reproduce the problem. In your case it should be pretty easy based on how often it has occured during the course of this thread.

Most likely re-imaging the machine and installing the drivers from the ppa would be the first line of attack. If that doesn't fix the problem, it most likely is hardware.

If re-imaging is first line of attack, I guess I might as well go ahead and try that here. It would lead to a shorter turn-around time for me if it works and less costs for you. If changing the mouse/keyboard doesn't help I'll try it out. If there's any special way you want me to do it, and in particular, if there's anything you want me to make sure I *don't* install, please let me know.

Then, we'd start swapping things out: power supply, graphics card, cpu, motherboard, etc...

Ok. If the above two things don't work lets go for it and you can swap out as much as is necessary to find and resolve the issue.

Of course, you'll be able to repeat the reimage, if you want, once you get it there, but if that's going to fix it I'd prefer to avoid sending it in.

If you have any concerns or suggestions about this please let me know.

Thanks. - corey

[thomasaaron]

I wasn't aware you sent a USB (non-wireless) keyboard. I have not received it yet if you have. I don't have one to try unless I go buy one or get the one you've sent. Can you please send a tracking number via email?

I didn't. I'm sorry, but I got you mixed up with Ed.Corcoran (he has the other wild dog post). It's getting hard to keep it all straight! At any rate, I sent him my only extra keyboard. If you can borrow one, though, it probably wouldn't hurt to eliminate wireless devices for testing. There is a fair chance they are playing a part in his issue, but it may not have any bearing on yours.

I don't have that program installed at all (just confirmed) and wasn't even aware of it until you mentioned it here.

Yeah, scratch that one.

Restore instructions are here...

You can download a CD image for Ubuntu here...
http://www.ubuntu.com/getubuntu/download
...I recommend using 9.04 64-bit.

Once you've downloaded the CD image, you will need to burn it to a CD. However, it must be burned as an *image* and not as a *data * CD. Otherwise, it will not be bootable. Instructions for this vary based on which operating system you are using to burn the CD. Instructions for various operating systems are here.
https://help.ubuntu.com/community/BurningIsoHowto

Here are instructions for re-imaging your machine...
http://knowledge76.com/index.php/Restoring_Your_System
...including instructions for downloading, installing and running your System76 driver.