Thanks for your advices,

The system is connected with a NAS server, the hardware connection was not changed since they were in solaris 8.

There are many coredumps under /var/crash/hostname, i've checked through these images by mdb, found CPU status and stack info are different each time, can you tell how can i parse the images of this kind of panic? The system is a cPCI processer card, with 650MHz sparc cpu.

A sample piece of messages around the panic is below:
Jul 31 19:19:51 vzss7sip2 xntpd[232]: [ID 774427 daemon.notice] time reset (step) 3.113106 s
Jul 31 19:20:32 vzss7sip2 fxp: [ID 786680 kern.notice] fxp1 : No response from Ethernet network : Link down -- cable problem?
Jul 31 19:24:16 vzss7sip2 unix: [ID 257829 kern.warning] WARNING: interrupt 0x7c0 at level 15 not serviced
Jul 31 19:24:16 vzss7sip2 unix: [ID 836849 kern.notice]
Jul 31 19:24:16 vzss7sip2 ^Mpanic[cpu0]/thread=2a100117cc0:
Jul 31 19:24:16 vzss7sip2 unix: [ID 534432 kern.notice] dispatcher invoked from high-level interrupt handler
Jul 31 19:24:16 vzss7sip2 unix: [ID 100000 kern.notice]
Jul 31 19:24:16 vzss7sip2 genunix: [ID 672855 kern.notice] syncing file systems...