XBMC randomly consumes system load (crash/hang)
#31
Ken,

I just noticed you seem to be using a fairly early version of XBMC (26018), which is from before even the pre-addon merge. I was thinking that this may be a merge version related issue, but apparently not. The reason for my thinking was that I cannot recall ever having this issue on the pre-addon SVN PPA.

So, I suppose my next experiment will be to go back to the 28276 build I have shelved and see if this issue can be reproduced there. If so, this may be an issue that is going deeper than XBMC itself.

It also seems possible that although our issues are similar, they are not directly related. There is also this thread going on which I think may be related: http://forum.xbmc.org/showthread.php?tid=77201

Fred
Reply
#32
Well, in my case, I believe my problem is very very likely hardware related. I came to this conclusion after doing the following:
  1. I booted to a raw XBMC Live disk to check if this was caused by some odd driver issue I might have introduced, or if it was an issue with my SVN builds. It still crashed. This led me to believe I was having a HW issue, so I started rolling back possible BIOS and HW changes I had made.
  2. I remembered I set AHCI on my SATA controller. Set this back to Native IDE... crashed.
  3. I had recently installed a second SATA card. Removed... Crashed.
  4. After removing the card, I noticed the area around video/south bridge was hotter than I expected. Inspected video card... fan not turning and feel stiff. I tried cleaning it, still not turning. Looks like a very very likely culprit.

I'm on my way to the store to try and find an adequate replacement fan. Card seems okay.

UPDATE: So far so good.

Fred
Reply
#33
webmosher Wrote:UPDATE: So far so good.

Can you keep us further updated within a few days to helps us making conclusions here?

regards
Reply
#34
Well, in my case, the problem seems to be completely resolved. It ended up that I just bought a new video card (a cheap GT210) for the time being. My system has stayed up almost 24 hours with no significant issues.

It seems that other issues mentioned in this post are using a variety of hardware. So, to try and isolate those types of issues, I would suggest grabbing gtkperf or just running glxgears for a few minutes to an hour to check if you are having GPU problems like mine. This basically takes the problem outside of XBMC's domain and helps to confirm its caused by other factors. These tools might not be stressing the GPU enough, but they might give you an idea if that is a direction to pursue.

You can also try doing some physical checks. If you open your case, and you note significant heat sources, then that might help to pinpoint possible issues. I'm not saying warm... the hardware will likely have a bit of heat, but if you inspect the system (while turned off) after a failure and there are components that are very hot to the touch, that's suspicious. Its also helpful to physically look to make sure fans are turning, etc. General areas to look: CPU & fan, this should be fairly cool. South bridge (usually has a sink or a fan)... if you have a fan, make sure its turning. It might be warm, but not hot to the touch. Lastly GPU... if you have an all-in-one MB, this should be a separate heat sink or might be built into the south bridge.

The lm-sensors package might give you a good idea how hot its running while you're in xbmc. It should be a good indicator of problems, but was not conclusive in my case. This is a decent guide for that: https://help.ubuntu.com/community/SensorInstallHowto

In the case of heat issues with hardware, this seems random because you are waiting for a threshhold that the heat must reach before the failure occurs. However, stressing the GPU or CPU to failure should be simple in these case: In the case of the GPU, if you use VDPAU, then simply find a high bit rate H264 movie and play it. If you have some sort of graphics benchmark, that also should work. There are some episodes in the Planet Earth series that have high rates, but they are not usually sustained. There may actually be test files online that may help with this. It should fail fairly quickly if there is a GPU issue. If you suspect CPU issues, then turn VDPAU off and play the same movie. Memory is also suspect to heating issues, but this usually only happens if you are over clocking. However, faulty RAM can cause a wide variety of weird and random problems.

From a software side, the key is reproducing the error in some generally expected way. If it seems to fail in some circumstances "randomly", try finding a situation where it always fails. Generally though, in my experience, random software crashes are fairly rare unless there is a weird memory leak. I should have realized this from the beginning of my testing. These random issues usually indicate a piece of hardware is flaking out.
Reply
#35
I'm a little further in my investigtions

http://forum.xbmc.org/showthread.php?p=5...post570759

rgs
Reply
#36
webmosher Wrote:Well, in my case, I believe my problem is very very likely hardware related.

I am starting to wonder about this as well. I usually use the box for XBMC only but recently received a link to a video on the web. While watching the video the system hung as well. This makes me wonder whether the problem is outside of XBMC.

I don't have an external video card; I only use the on-board graphics card. And I am not really sure what to do to test the problems further. If it is a hardware/graphics problem, then why does the hard-drive light turn on and the system come under so much load? Is there a good way to test whether the problem is with the graphics card? I suppose this means buying a graphics card. Any suggestions on what to look for in order to keep costs and wattage/heat down?

Thanks.
Home Entertainment PC: XBMC running on Ubuntu 10.04
AMD 9950 Phenom Quad on ASUS M3N-H/HDMI motherboard with 4GB RAM and two SATAII drives.
Reply
#37
kenm_uk Wrote:I am starting to wonder about this as well. I usually use the box for XBMC only but recently received a link to a video on the web. While watching the video the system hung as well. This makes me wonder whether the problem is outside of XBMC.

I don't have an external video card; I only use the on-board graphics card. And I am not really sure what to do to test the problems further. If it is a hardware/graphics problem, then why does the hard-drive light turn on and the system come under so much load? Is there a good way to test whether the problem is with the graphics card? I suppose this means buying a graphics card. Any suggestions on what to look for in order to keep costs and wattage/heat down?

Thanks.

in the first place I would suggest to read the threads well, and then determine comparing on the posted problems what your problem exactly is.

to me, the hardware temperature drift was absolutely not the problem.
Check the link I've sent.
If you read well, you'll see in this thread also several types of problematic behaviour of the xbmc box under certain conditions.

It is your task to find out what the behaviour of your problem exactly is.
You fall under the same conditions as ours? -> contribute to the thread
if not? -> start a new thread

What is the behaviour of your sysem exactly?
idle -> hangup?
play video -> hangup or crash?

post this to the forum, it is more useful.

Good luck!
(more exact described problem/behaviour lead to easier problem resolution)
Reply
#38
I've been having the same total system hangs on Windows 7 recently. (On a HP DV5 notebook with ATI Mobilty Radeon HD 3200 integrated Graphics).

I'm pretty sure I've identified my particular problem as being related to power states. See my post here:


Hope that helps anyone else!
Reply
#39
Update: Over the last few evenings, when XBMC appears to hang I have just let it be. After anywhere between 5-15 minutes XBMC comes back to life again.

Again, for these 5-15 minutes the hard-drive light is nearly constantly on and the XBMC screen is frozen. It is very very difficult/slow to SSH into the box or access the Samba file share.

I wish I know what was happening but it is interesting to note that it is obviously doing something and just needs some time to finish. But it is very annoying if after watching a video I have to wait 15 minutes before the box becomes usable again.

Does anyone have an idea what is going on? Is XBMC updating its databases or scanning for new media?
Home Entertainment PC: XBMC running on Ubuntu 10.04
AMD 9950 Phenom Quad on ASUS M3N-H/HDMI motherboard with 4GB RAM and two SATAII drives.
Reply
#40
also an update from my side:
the problems remain...

I've installed a new xbmc's latest version in an attempt to solve this, this did not help at all.

Right now, I'm monitoring the box with top and atop via ssh.

HDD light is on
CPU takes not more then normal, 2%-3%

dunno anymore what's helpin' ... Sad
Reply
#41
So, I'm into it a bit deeper now...
  • Anyone of you guys having a Asus mobo? Mine: P5N7A-VM
  • The problem I've seen is to be found in the /var/log/syslog
    At startup time, by default after vanilla ubuntu install: (cleaned out log below)
    Quote:ACPI: I/O resource nForce2_smbus [0x4d00-0x4d3f] conflicts with ACPI region SM00 [0x4d00-0x4d3f]
    ACPI: If an ACPI driver is available for this device, you should use it instead of the native driver
    nForce2_smbus 0000:00:03.2: Error probing SMB1.

    And further when the system "hangs" (it is in fact a crash)
    Quote:ata2.00: exception Emask 0x10 SAct 0x0 SErr 0x400000 action 0x6 frozen
    ata2.00: irq_stat 0x08000000, interface fatal error
    ata2: SError: { Handshk }
    ata2.00: cmd a0/00:00:00:08:00/00:00:00:00:00/a0 tag 0 pio 16392 in
    cdb 4a 01 00 00 10 00 00 00 08 00 00 00 00 00 00 00
    res 50/00:03:00:14:00/00:00:00:00:00/a0 Emask 0x10 (ATA bus error)
    ata2.00: status: { DRDY }
    ata2: hard resetting link
    ata2: softreset failed (device not ready)
    ata2: hard resetting link
    ata2: softreset failed (device not ready)
    ata2: hard resetting link
    ata2: link is slow to respond, please be patient (ready=0)
    ata2: softreset failed (device not ready)
    ata2: limiting SATA link speed to 1.5 Gbps
    ata2: hard resetting link
    INFO: task hald-addon-stor:1189 blocked for more than 120 seconds.
    "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
    hald-addon-st D 0000000000000000 0 1189 958 0x00000000
    ffff8800c3441b78 0000000000000082 ffff8800c3441b08 0000000000015880
    ffff8800c48b83b0 0000000000015880 0000000000015880 0000000000015880
    0000000000015880 ffff8800c48b83b0 0000000000015880 0000000000015880
    Call Trace:
    [<ffffffff8133d8f5>] scsi_block_when_processing_errors+0xd5/0x100
    [<ffffffff81078570>] ? autoremove_wake_function+0x0/0x40
    [<ffffffff8134e1b1>] sr_open+0x11/0x20
    [<ffffffff813967cd>] cdrom_open+0x12d/0x200
    [<ffffffff8126beb9>] ? get_disk+0x49/0x80
    [<ffffffff8152b9a9>] ? _spin_lock+0x9/0x10
    [<ffffffff8127850a>] ? kobject_get+0x1a/0x30
    [<ffffffff8134e394>] sr_block_open+0x74/0xd0
    [<ffffffff81278367>] ? kobject_put+0x27/0x60
    [<ffffffff8114d1f2>] __blkdev_get+0x1f2/0x3d0
    [<ffffffff8114d3e0>] ? blkdev_open+0x0/0xc0
    [<ffffffff8114d3db>] blkdev_get+0xb/0x10
    [<ffffffff8114d44c>] blkdev_open+0x6c/0xc0
    [<ffffffff8111d5d6>] __dentry_open+0xe6/0x2e0
    [<ffffffff8111d8d7>] nameidata_to_filp+0x57/0x70
    [<ffffffff8112cf3a>] do_filp_open+0x2ca/0xad0
    [<ffffffff8114ce3e>] ? __blkdev_put+0x7e/0x1a0
    [<ffffffff811378c2>] ? alloc_fd+0x102/0x150
    [<ffffffff8111d3a4>] do_sys_open+0x64/0x160
    [<ffffffff8111d4cb>] sys_open+0x1b/0x20
    [<ffffffff81012042>] system_call_fastpath+0x16/0x1b
  • This problem occurs with any type of hdd (sata/ata) I install Karmic & xbmc on.
    Maverick gives also the first problem, xbmc not tried as yet.
  • The problem occurs after playing a movie or after a long play shoutcast or so.
  • The moment of xbmc-hang/crash the 2nd quote is logged.

Anyone having the same crash behaviour ?

rgds, Jo
Reply
#42
Hi Everyone,

I've come to a current solution and reason of hanging.
As you can see in the call trace, a "cdrom_open" is mentionned.
Leaving out the DVDdrive solved my problem.
Now I need to find out how to have it connected and not causing a crash Smile

I'll keep the thread posted if it appears to be not the real reason.
Reply
#43
After a number of months I think what happens is...
everytime a new file is added to a folder (which I have named new_media) and I try to access that folder in XBMC, the hard drive light starts up as if XBMC is scanning all the new media in that folder for changes. It hangs XBMC for 3-10 minutes and then resumes operating as normal.
I don't know how to provide more debugging options but it might be useful to the developers to have an idea what is going on in this situation. If I can help please let me know how.
Home Entertainment PC: XBMC running on Ubuntu 10.04
AMD 9950 Phenom Quad on ASUS M3N-H/HDMI motherboard with 4GB RAM and two SATAII drives.
Reply

Logout Mark Read Team Forum Stats Members Help
XBMC randomly consumes system load (crash/hang)0