SUMMARY: what does mean such error message?

From: Grzegorz Bakalarski <G.Bakalarski_at_icm.edu.pl>
Date: Tue Aug 05 2003 - 10:06:20 EDT
Hello ALL

Sorry for late summary, but seem it is not serious. So this is rather 
in order to be archived for future:


Thanks to Mike.Salehi@kodak.com Christopher.Wilkinson@gfk.de 
hans.schaechl@kippdata.de svaka@sitelite.com  and others.

All were pretty sure this is not problematic message if it appears rarely ...

Kind regards,

Grzegorz

8888888888888888 Original Query 88888888888888888888888888888888888
Maybe someone could state what does mean a following error:

Jul 28 13:09:03 zatoka SUNW,UltraSPARC-III+: [ID 762062 kern.info] NOTICE: [AFT0] WDC Event
+detected by CPU6 at TL=0, errID 0x000749f6.f584087c
Jul 28 13:09:03 zatoka SUNW,UltraSPARC-III+: [ID 796924 kern.info] [AFT0] errID 0x000749f6.f584087c
+Data Bit 102 was in error and corrected

????
Look like not severe, but might it be a sympthom of something
more dangerous that will come soon ???

machine: 1year old V880, solaris 9


88888888888 Answer 1

AFT0  messages are usually memory error, but sun requires 3 persistent
(this is not clear in this case) cases.


888888888  Answer 2

ECC
_____________________________



888888888 Answer 3

the "WDC event" is a correctable ECC error detected by one
of the CPUs of your system. Since you use a V880 with Ultra-
SPARC III+ CPUs the system will try to correct level 2 SRAM
ECC errors either by hardware or software.
WD= writeback data C= correctable
So this is a good thing! Otherwise with older CPUs without ECC
error detection and correction your system would have panic'ed.

If the errors do not increase in number or always happen on the
same CPU module there's no reason to think of a faulty piece of
hardware. 


888888888 Answer 4

AFT0 errors are correctible memory errors. These are konly kernel informational messages. No action
+necessary.
If there are AFt1, AFT2 errors repeating several times in a 24 h period(which are non correctible),
+get it evaluated by SUN.   
That could indicate some CPU/memory board related problems which might need repalcement.


8888888888888888888888888888888 EOT 88888888888888888888888888888
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers
Received on Tue Aug 5 10:10:22 2003

This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:43:17 EST