I'm having an odd problem with the new myth system I built. Occassionally I will come home and the system is completely locked up (still in frontend). I can't restart x, SSH in, or anything so I have to hit the reset button to get it back up.

Perhaps unrelated or not, the backend isn't always starting on boot.. I have to ALT-X and sudo mythbackend & to get it going. After that everything works fine.

I'm still green when it comes to linux, I know there should be some way to tell from log files or something similar what is causing the lockups, but I don't even know where to begin.. Can someone point me in the right direction? I'm running R5E50, ASUS M2N32-SLI motherboard with a AMD Athlon 64 X2 3600+ Windsor 2GHz processor, Geforce 7950GT video card.

Thanks in advance for your help.

Michael

Unfortunately I can't help you on the freezing side, except to suggest you run memtest for a day and see if you get any errors. Generally hard freezes like that are caused by hardware either failing or overheating

The backend not starting problem is a known issue having to do with the script that starts it. If you search the forum someone posted a fix that seems to have cleared it up for me. In the meantime, if you press alt-3 it should restart the backend (alt-2 stops it).

Hi,

I have had that happen also when using a large serial drive. IDE has been working ok for me.

Mike

I do have a 750GB SATA drive.

HI,

I have not had the chance to try running KM off an ide and use the serial just as the /myth storage. Just a thought. (Do a backup first!)

Mike

I'm having the same problem, using a 320GB IDE hard drive, on a P4 2.26Ghz machine, 512MB ram, with 2x Hauppauge PVR500's as a back end machine, and it has been locking up constantly. Running windows XP no problems, running FreeBSD 6.2 no problems, running regular Knoppix off of a CD no problems. Only has been happening with Knoppmyth. Also I some times get an odd problem with the PVR500's where the recording gets a little jumbled, say i watch the same show on my TV and also even live on the box, the tv always is nice and stable, and on the box it will sometimes jitter the video. Almost like it's not getting it's interrupts serviced enough. And that is only with one pvr500 doing anything, recording on only one encoder, the other idel and the other card also idle. Anyone have any ideas?

With Lockups there are so many reasons for this to be happening it is had to discern what the exact root cause of the problem is. Here are some tips to try and find where the problem may lay.

1. Physically inspect the Mainboard
- What you are looking for is busted or popped capacitors, this has plagued many Mainboards for sometime and the reason a lot of manufacturers are moving towards using solid capacitors. Capacitors are the tall things generally around the CPU socket and have a cross on the silver top. If the top is raised or you can see a light brown powder seeping out the top you know your in trouble.

2. Memtest
- This test will test your memory (RAM) and it is important to let this test run for some time as the problem may be caused by components when under stress, I would suggest at the very minimum of 3 passes.

3. Heat
- This is sometimes hard to determine, the simpliest way I can suggest to see if this is the issue, is as soon as the computer locks up, reboot the computer and go into your BIOS. In there you will have something like "PC Health" which will list the temperatures inside the system and fan speeds. You could find that the system is overheating or a fan my be under performing.

4. Power supplier
- This is sometimes the least thought about component and is often neglected as a $40(AU) case can include a power supply thus why would you purchase another one. Think of it this way, if you pay $40(AU) for a case and PSU, why would you spend btn $80(AU) - $200(AU) on a decent one when I already have one. Power is essential to the stability of the system, without clean, reliable power the system is not worth a pinch of poop.
To look at this issue, in your BIOS as in suggestion 3, check your "PC Health", there should be voltages listed, check they are very close to what they should be. Now this is only an indication as the system is not under load and that is where the problem could be. If possible unplug unnecessary components, eg other HDDs, stuff like that.

5. Failing HDD
- You could test the Hard Disk Drive with the manufacturers software to ensure that it is not failing. You can download the software from the their website or you can use this disc Ultimate Boot CD which contains a number of utilities for testing system components including a number of drive manufacturer tools.

In saying all that it could be the software, but unlikely due to the large install base and few user reporting similar problems. If you have a particular problem don't hesitate to post a reply or drop me an email or PM.

_________________
Girkers

Hi,

Those are good tips especially for home built / older hardware. The boxes I had the issues were Dells, One fresh out of the box and one only slightly older with new 500 gig drive. I backed down to R5C7 and they run like rocks.

I'm suspecting some "energy saving" mode on the new large drives but with out some kind of trail it is tough to travel. Could even be some simular control in the kernel. More questions then answers.

Mike

I'm still having the random lockups, it just happened and I was logged into the computer over ssh.. I saw this was spit out in the SSH terminal. I'm still very new to linux so I'm not really sure what it all means.. Can someone help decipher?

I just did a search on Google in regards to this error and from what I read it may have to do with multicore processing and the kernel. You could try turn off Hyperthreading (Intel) or run the processor as a single core (if using dual core) to see if this eleminates you problem

I also would like to ask, did you do any of the tests I suggested?

_________________
Girkers

I have an AMD processor so turning off hyperthreading is not an option. I dug around a little in the BIOS and didn't see an option to switch to single core, where is this normally?

I did run memstest for 2 passes with no errors. According to RRDTool the system and CPU temperature has been very constant (the reboots show up as breaks in the graph and there's no spike before the break).

Here's my complete system:

ASUS M2N32-SLI Deluxe Wireless Edition

AMD Athlon 64 X2 3600+ Windsor 2.0GHz Socket AM2 Processor Model ADO3600CUBOX - Retail7200 RPM SATA 3.0Gb/s Hard Drive - OEM

XFX PVT71JYHE9 GeForce 7950GT 512MB 256-bit GDDR3 PCI Express x16 HDCP ExTreme Edition Video Card

SILVERSTONE LASCALA 10 CS-SST-LC10B Black Aluminum front panel, 1.0mm SECC body ATX Media Center Computer Case

Thermaltake TR2 W0070RUC ATX 430W Power Supply - Retail

Seagate Barracuda 7200.10 ST3750640AS 750GB

Kingston HyperX 2GB (2 x 1GB) 240-Pin DDR2 SDRAM DDR2 800 (PC2 6400) Dual Channel Kit Desktop Memory Model KHX6400D2K2/2G - Retail

What tuner card(s) are you using?

Marc

Oops..forgot to include those.. I pulled them from my last mythtv system (which had no lockups)

Hauppauge PVR-150 and Hauppauge PVR-250 (that's what the box said, but it's a bttv chipset)

I had some stability issues quite a while back that I resolved by changing some of the settings in /etc/lilo.conf. I modified the line that reads "append=" to look like this:
I believe it was the "noapic" that was key for me. I don't know if this will help you, but if you decide to try it, you first edit the file, then execute "lilo -v", then reboot, all as root.

Marc

Thanks for the tip Marc, I just made the change. I did notice on reboot that it seems to be a lot faster. The typical boot text was screaming by and the frontend seemed to load faster. Not sure what this means, but I'm not complaining!