raob@mullian.ee.mu.OZ.AU (richard oxbrow) (03/05/90)
Our sun4/390 (32Mb ram,2 ethernet cards,IPI disk) has started to crash
with great reguarlarity. During a normal week day it is -* expected *-
that the machine will crash at about 10 minutes to the hour, with the
following message
[[Ed's Note: It looks like the Lance ethernet card might be causing the
problem; although the problem described in v9n63 is not the same as yours
(it doesn't crash the Sun), it might be worth reading. -bdg]]
Feb 19 15:50:47 gondwana vmunix: tmd=ff06e258, flags=81, chip=ff01a000,csr=a2f3
Feb 19 15:50:47 gondwana vmunix: panic: TINT but buffer owned by LANCE
Feb 19 15:50:47 gondwana vmunix: syncing file systems... [27] 1 [27] [24] [18]
[11] [1] done
Feb 19 15:50:47 gondwana vmunix:
Feb 19 15:50:47 gondwana vmunix: dumping to vp ff013c4c, offset 66122
Feb 19 15:50:47 gondwana vmunix: SunOS Release 4.0.3.SPARCserver.390
(ECR.4.390)
#2: Fri Jan 5 15:53:33 EST 1990
Feb 19 15:50:47 gondwana vmunix: Copyright (c) 1989 by Sun Microsystems,
Inc.
Feb 19 15:50:47 gondwana vmunix: mem = 32768K (0x2000000)
Feb 19 15:50:47 gondwana vmunix: avail mem = 31956992
The 4/390 is configured to run as a gateway between a network of 60 sun3
workstations and the rest of university campus.
Has anyone had similar problems ? Does a patch exist for this problem
already ?,
At the moment we are sending sun the core files from the crashes.
Thanks ..
richard ..
#ps. The problem would would appear due to brittle ethernet code in the
4/390, since the problem appears to go away when we disconnect from the
university network. No other suns (3/{50,60,110,140} 4/{60,110,260}) on
campus would appear to have the same problems.