radmon.org testing new server!

James_Gy wrote:

mw0uzo wrote:

All I need next is the mobo to have bad caps...

Don't tempt fate

Luckily this week has been a slow work week, so had plenty of time to get this damn thing sorted out. Seems to be stable again now. I can only pray that it stays alive overnight.

Bah! This time 'BUG: Soft CPU lockup error'
This time is was only at 8:10am, about an hour ago.

I am going to remove the CPU cooler, clean and reapply paste and set a slight underclock.

Ok new paste installed, the old stuff didn't look like it had aged well at all. Cooler with copper core and more metal. Underclock to 150Mhz from 166Mhz. If this isn't stable now, then I will have to change the mobo. I have an Intel board left and a few P4 and celeron D processors and coolers in a drawer, so not at the end of the road yet. Luckily linux doesn't complain much about drastic changes of hardware

Might have found another source of problem, the IEC power cable was loose and might have a questionalble contact. Moving it around caused a reboot!

Server has been up all day without issue. Hopefully it will survive the night? Please?

If not, I have prepared an Intel mobo with P4 2.4Ghz Northwood. I tested the RAM stick that came out of the server and it showed problems, which disappeared after cleaning the edges with contact cleaner. Must have been a thin layer of oxide leached through the gold causing bad/intermittent contact. That should have shown up in memtest but it didn't. I probably didnt leave it going long enough. The ethernet interface has not been recognised (bah) so I will have to fix that if possible. Fun.