[VU+ Duo2] Box starting to crash after about 24 Hours use

[lincsat]

Starting to get the box crashing after about 24 Hours of use, it crashes on every auto enigma restart needing a power cycle to fix. It looks like the appropriate part of the crash log relates to the softcam check reporting no space on device yet there is plenty of space in RAM, Flash and on USB + HDD

Traceback (most recent call last):
  File "/usr/lib/enigma2/python/Plugins/SystemPlugins/ViX/SoftcamManager.py", line 551, in softcam_check
IOError: [Errno 28] No space left on device
< 71356.034766> [ePyObject] (CallObject(<bound method SoftcamAutoPoller.softcam_check of <Plugins.SystemPlugins.ViX.SoftcamManager.SoftcamAutoPoller instance at 0x6009ca58>>,()) failed)

Full log attached

View attachment enigma2_crash_1434109181.log

 

[judge]

Anything else filling up /var/volatile/ at the time of the crash?

 

[DaMacFunkin]

Have you got debug logs or something else enabled that could be filling your flash?
As a side note my first duo2 had to go back because of a memory problem that seemed to arise after a certain amount of use (I always said it was a faulty joint) the unit became increasingly flaky till it wouldn't boot anymore (vu+ where unable to fix it and issued me a new unit).

 

[lincsat]

Anything else filling up /var/volatile/ at the time of the crash?

Not that I'm aware of

Have you got debug logs or something else enabled that could be filling your flash?

As stated, the image is stating plenty of free RAM, Flash and an unused swap file

Could a dev say what's at - "/usr/lib/enigma2/python/Plugins/SystemPlugins/ViX/SoftcamManager.py", line 551

 

[Trial]

Hi,
closing file /tmp/cam.check.log when timer check is enabled the check writes information into this file.

ciao

 

[judge]

As stated, the image is stating plenty of free RAM, Flash and an unused swap file

There might be plenty of free flash but /var/volatile/ is it's own partition on flash & it looks like that's what's full.
Also, why are you using a swapfile on a Duo2?

 

[lincsat]

It only started happening a few days ago, so has something changed with the cam check logging? How to you set a max size for the cam check log?

I always set a swapfile and let the system manage it.

 

[judge]

so has something changed with the cam check logging?

Not that I know of, but anything else could also be logging to /tmp including all sorts of plugins.

I always set a swapfile and let the system manage it.

There really is no need for a swap file on a Duo2. I'd fresh flash without creating one in order to start debugging your issue.

 

[abu baniaz]

Meant to be writing to /tmp/ directory
https://github.com/OpenViX/vix-core/blob/master/src/SoftcamManager.py#L551

File has not changed since 14/4/2014

if config.softcammanager.softcamtimerenabled.value:
    #             print "[SoftcamManager] Timer Check Enabled"
    output = open('/tmp/cam.check.log', 'a')
    now = datetime.now()
    output.write(now.strftime("%Y-%m-%d %H:%M") + ": Timer Check Enabled\n")
    output.close()
    self.timer.startLongTimer(config.softcammanager.softcamtimer.value * 60)
else:
    output = open('/tmp/cam.check.log', 'a')
    now = datetime.now()
    output.write(now.strftime("%Y-%m-%d %H:%M") + ": Timer Check Disabled\n")
    output.close()
    #             print "[SoftcamManager] Timer Check Disabled"
    softcamautopoller.stop()

 

[DaMacFunkin]

I don't have softcam timer check enabled on my solo2 and duo2, since I moved to oscam i don't think it has ever froze.

 

[pembo]

Have you turned on lcd4linux logging?
have you tried logging in and simply checking whats using the space in /tmp?

du -h /tmp/*

Will show you the space used within /tmp

 

[lincsat]

I had not checked /tmp/ while it was crashing as I didn't know at the time it was the "device" that was being written to and after a reboot it's empty.

I've just checked and LCD4Linux logging is off. The camcheck log had reached 18k in 24 Hours, I have now disabled the camcheck as well, it had been left as default.

 

[judge]

The camcheck log has never filled my /tmp/ on all boxes.
I'd suspect it's something else writing to /tmp/

 

[lincsat]

Up until 4 Days ago, I'd never had the problem. Then happened twice in just over 48 Hours and been OK since. I'm keeping an eye on /tmp/ for anything else in there. How much space is allowed for /tmp/ in a Duo2? I suspect it's a few meg.

 

[judge]

Up until 4 Days ago, I'd never had the problem. Then happened twice in just over 48 Hours and been OK since. I'm keeping an eye on /tmp/ for anything else in there. How much space is allowed for /tmp/ in a Duo2? I suspect it's a few meg.

df -h

in a telnet session will show you the size of /var/volatile/

 

[lincsat]

Mmmn - size - 605.5M, Used - 396.4K, Available - 605.1M.

There is no way a log file is going to take up over 605 Meg. I'll keep an eye on it but over 32 Hours since last reboot, so maybe it's cleared itself, whatever it was.

 

[john doe]

605m is a big file

 

[lincsat]

I think I've found the offender but no idea what caused it. /var/volatile/log/log.nmbd was 605 Meg! It's too big to open in a text editer, so juse deleted it.

It's something to do with Samba, so I've removed samba from the box and re-installed and so far it's at a healthy 149 B

 

[abu baniaz]

Perhaps related to windows share issue you had?

 

[judge]

I think I've found the offender but no idea what caused it. /var/volatile/log/log.nmbd was 605 Meg! It's too big to open in a text editer, so juse deleted it.

It's something to do with Samba, so I've removed samba from the box and re-installed and so far it's at a healthy 149 B

Using an old backup?
Edited any config files on the box?
Something on your network making flakey nmbd requests?