Dramatic Connection Table increase after SmartDefense Update

[tyoung]

I was running NGX R60 HFA03 on two Nokia IP1220 gateways (redundant via Nokia VRRP) and a Sun Sparc Ultra-60 with SunOS 5.9 serving a mgmnt server. I upgraded to SmartDefense and Web Intelligence version 591080302 one night but didn't push the policy or do a revision control (ouch). The next day I created an object and added it to an existing rule and pushed the policy. Within an hour the master gateway was refusing new connections and the connection table had overrun its 25K limit. I increased it to 50K and it overran that. I increased it to 100K where it remains. The connection table runs into the upper 70K's every day now. 90% of the connections are related to two redundant forward MS ISA proxy servers. But they have always been there. I deactivated SmartDefense and Web Intelligence protection but that didn't help so I re-enabled them. I've upgraded to HFA_05 which didn't help. Tomorrow I'm going to deactivated SmartDefense and Web Intelligence protections again. I've had encryption cards in the 1220's and have had Nokia's SecureXL enabled for some time. But I do not have Sequence Verifier enabled in SmartDefense or Nokia's Voyager. Anyone know what may be happening, if so, please do advise?