nopers i can debunk that easily. on the 2 gig machine login as root and issue swapoff -a and let it go for 24 hours..you'll see the box start getting oom errors and most likely more than a few outages as packet processing come to a halt as the system is simply too starved for ram to continue.
You should NEVER have to reboot a properly tuned Linux box(or windows server 2k3 and above for that matter) to "fix an issue". That's not a fix and only covers it up. The system isn't releasing the swap space because it know it is going to need it again. That along with Sopho's predilection to NOT properly utilize file cache(aka releasing that) as well leads to the system using swap like that.
nopers i can debunk that easily. on the 2 gig machine login as root and issue swapoff -a and let it go for 24 hours..you'll see the box start getting oom errors and most likely more than a few outages as packet processing come to a halt as the system is simply too starved for ram to continue.
You should NEVER have to reboot a properly tuned Linux box(or windows server 2k3 and above for that matter) to "fix an issue". That's not a fix and only covers it up. The system isn't releasing the swap space because it know it is going to need it again. That along with Sopho's predilection to NOT properly utilize file cache(aka releasing that) as well leads to the system using swap like that.
