My messages log is full of ip_conntrack: table full, dropping packet
ip_conntrack_max is set at 65536
The server is running Redhat Enterprise 4 release 4 and fully updated. It has 2 GB of memory and uses 1 GB in production.
It has 4 interfaces and is used a router and a firewall with iptables.
What is the limit on ip_conntrack_max? Can I increase it more? How can I resolve teh problem of the table being full?
Thanks,
Jay

Yoloits wrote:

What is the limit on ip_conntrack_max? Can I increase it more? How can I resolve teh problem of the table being full?

You can increase it easily (echo a-huge-number
/), but a better way using
the almost-unknown hashsize option of ip_conntrack module:

We had the same issue and the solutions above helped.