I think you need to rule out the kvm first.
It could be overheating, but modern cpus should be stable to at least 60 or
70 deg c as measured on silicon, maybe higher.
The psu seem less likely to me, but I've never had one fail, other than
completely, but have had every other thing fail over the years and produce
odd errors.
Ram can be tested easily, the rest is a bit harder to test. Cpus normally
only overheat but other components on a motherboard can be more finicky.
It is worth saying I have had cpu overheating issues from old thermal
paste. Removing the cpu cooler, cleaning it and reseating with new, quality
thermal paste solves that.
Anton
On 26 Jan 2013 23:05, "Leszek Kobiernicki 1" <l.kobiernicki@???>
wrote:
> On 26/01/13 14:30, Rob Malpass wrote:
>
> Hi all****
>
> ** **
>
> My dad (an electronics engineer of 40 years experience) once told me
> “intermittent faults are a swine to fix” – and never truer words were
> spoken. Could you all please take a look at my logic before I condemn
> certain parts of this failing box to the bin?****
>
> ** **
>
> 2008 vintage 64-bit 4Gig RAM machine running Ubuntu 12 fine for the past
> month – 28 days uptime as my media server and no problems at all. Today I
> switched the kvm box controlling it to control another machine (no
> disconnection, just a flick of a switch) and the box powered down! For
> some reason whenever the bad box powers down, it needs to be physically
> unplugged from the mains before it will come back up – this particularly
> baffles me. Brassed off with this, I started to look into what causes
> random shutdowns (it’s not the first time it’s done this but as I say it’s
> been fine for 28 days). I left it running for 20 minutes in BIOS to check
> the CPU temp was fine – and it was at 44 degrees C max. Rebooted to see
> what dmesg might say but – post BIOS but before boot – it shut itself
> down. And at that point my anglo saxon became taboo for the good folks on
> this list!****
>
> ** **
>
> I’m now stumped and, as the machine is 2008 vintage (though good for its
> time) it might be time to upgrade – but I’m wondering which bits I can
> salvage. I’m thinking the following are ok to reuse: case, optical drive,
> hdd whereas any of the mobo, RAM and certainly PSU could well be the
> trouble and it’s almost impossible to test which.****
>
> ** **
>
> Can anyone offer another diagnosis / treatment for these symptoms? I
> might be inclined to buy a new PSU and see what happens swapping that
> before any more major surgery but *is there really anything I’ve missed
> as regards seemingly random shutdowns*? Incidentally when I say shutdown
> – I’m talking immediate power down – not the OS executing a halt command.*
> ***
>
> ** **
>
> Cheers****
>
> Rob****
>
> ** **
>
> Cheers****
>
> ROb****
>
>
> I'd put money on it, that you've a heat problem
>
> My boxes are never allowed to go beyond 36 degrees, before I power off
>
> They all have 2 additional fans - besides the ones they came with - to
> ensure component life isn't shortened
>
> BIOS powers off a box which goes beyond critical temperature: the manual
> should tell you what the manufacturer considers that to be
>
> Opinions seem to vary quite a bit on this point .. Wonder what others
> think about this .. ACRONIS Disc Manager states it is up at 50 degrees !
>
> A short-term fix, would be to take the side off, & put a small desk fan
> next to it, blowing cold air into the box, & help dissipate whatever is
> causing the pb.
>
> L
> --
>
> --
> Please post to: Hampshire@???
> Web Interface: https://mailman.lug.org.uk/mailman/listinfo/hampshire
> LUG URL: http://www.hantslug.org.uk
> --------------------------------------------------------------
>
--
Please post to: Hampshire@???
Web Interface:
https://mailman.lug.org.uk/mailman/listinfo/hampshire
LUG URL:
http://www.hantslug.org.uk
--------------------------------------------------------------
