One of my customers has a 220 (V7.511) with NetSec & WebSec including site-to-site VPNs to ten 120s (V7.511) in separate locations.
About 8 users in the central site and two users in each remote site surf occasionally via the HTTP/S Proxy in the 220. Wireless guests in each remote location are on a subnet that is not a part of the VPN tunnel; they surf directly via a Firewall rule and masquerading via an Additional Address, so there's no impact on the load of the 220 at HQ.
After 100 days of uptime for the 220 in HQ, swap had grown to over 40% and RAM use continuously ran in the mid-to-high 80s. Occasionally, heavy swapping drove CPU to 100%, though rarely did any sluggishness cause users to complain.
My first steps were to change to single-scan for AV and to ask the Admin to un-select items in Intrusion Prevent that he knew he didn't need.
A few days later, I learned that the problem continued unabated, so I glanced at 'Support >> Advanced' and noticed that the middleware daemon was using 18.6% of RAM - not good!
I had the admin reboot last night. This morning, I see the daemon at 6%, RAM at 72% and no swap space was needed for the database rollup last night - swap is at zero. Problem solved for now.
I didn't find this in the V7 KIL. Has anyone else noticed this phenomenon?
Cheers - Bob
This thread was automatically locked due to age.