We see that problem on Solaris 10u9. (varies in some sections)
fctl: [ID 517869 kern.warning] WARNING: fp(0)::6ad4c0 NS failure pkt state=d reason=9, expln=1, NSCMD=0100, NSRSP=8001
Problem is seen in „dmesg“ or /var/adm/messages after a „luxadm -e port“. In june oracle tell us „Works as designed“. Cant decalre it deeper, but the reason are some fc ports that are online in our SAN but not zoned. If we work as example on a M9000 domain with 6 HBAs (12 ports), 8 of them are used and 4 are free for whatever. Nobody put that 4 ports into a dummy zone. Now, if you do „luxadm -e port“ by hand or with the „explorer“ (SUN-Explorer does that automaticly) you get that failure in your messages. Today we found a solution -> Patch 145957-06 solved that issue. Oracle said „To resolve/filter out….“. Unfortunately you have to reboot that server, guess thats the biggest problem.