[SunHELP] Can u tell me whether the hardware is faulty ??

[Steve Pribyl]

I have heard that sun ultra cpu memory can't catch double bit parity
errors. This can cause undetected data corruption and will cause the box
to crash at some point.

However I don't know what ultra cpu's this applies to.  You should ask sun
about this.  They do have a fix.

The cause for these cases would really be sun spots and solar storms.

Steve Pribyl
spribyl at enteract.com
http://www.enteract.com/~spribyl

On Tue, 30 Jan 2001, Nenno, Tim wrote:

> We've had the same problem with a 5149-08 440Mhz IIi on an Ultra 10. That
> same panic message, plus lots of other CPU-related freezes/crashes/reboots.
> The machine would stay up anywhere from a few minutes to a week before
> failing.
>  
> The CPU's been replaced four times since November. One of them was DOA. The
> third was a week ago, the fourth yesterday. Yesterday's engineer replaced
> the system board, too, on the off-chance that a glitch there might be
> triggering the CPU problem.
>  
> In the course of all that, he tested the power supply, says he found a
> problem there that might be causing the CPU problem, and so replaced the
> power supply.
>  
> So, if the machine holds up, we won't know what the source of the problem
> actually was. (Not that I'll want to know it was the power supply all
> along....)
> 
> I have 220R with 2*450Mhz processors and 1Gb of Memeory..OS is Solaris-7
> Yesterday the system crashed with a core dump and i got the following
> messages
> logged in /var/adm/messages:::
>  
> unix: panic[cpu0]/thread=300002216a0:
> 
> CPU0 Ecache Writeback Data Parity Error: AFSR 0x00000000.00800004 AFAR
> 0x000001fe.01800800
>  
> Savecore: reboot after panic: CPU0 Ecache Writeback Data Parity Error: AFSR
> 0x00000000.00800004 AFAR 0x000001fe.01800800
>  
> Could you possibly tell me what went wrong ???
> 
>