I have one problem on two servers that i use as routers (two houses, two completely separate networks) , so i'm 110% sure it is something in my configuration. The issue happens without the iptables rules even loaded, so I'm 99% sure it does not have to do with iptables.
when i first turn on the routers, i ping google from it and it is under 20ms... but then if i do it again a couple minutes later, itll be 100ms+ until i reset the interface connected to the internet (ifdown eth1 && ifup eth1). The ping will go back down to 20, but a couple minutes later itll be back up to 100ms again.
configuration: my LAN router running debian:
eth0: LAN
eth1: internet
current kernel routing table:
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
localnet * 255.255.255.0 U 0 0 0 eth0
69.121.96.0 * 255.255.240.0 U 0 0 0 eth1
default ool-45796001.dy 0.0.0.0 UG 0 0 0 eth1
let me know any other information you need, I am desperate to resolve this as it is causing performance issues on both networks.
and after a couple minutes on a ssh connection into the box (from over the internet) it gets really laggy and then locks up... but then a new connection can be opened, act fine, then do the same exact thing.
another issue:
i have a webserver running on 192.168.1.5 on port 6080. When i request
http://domain:6080 from within the same LAN as the webserver, I get connection refused. When I connect to
http://192.168.1.5:6080 within the LAN, it works perfectly.
When someone outside the network requests
http://domain:6080, it works fine.
Again, i suspect this has to do with the kernel routing tables in my router.
edit: i think the difference in google pings may just be shifting routes chosen over the internet, for load balancing. I'm more concernde now with the laggy/frozen ssh connections and the webserver/local server access