CheckPoint: ISP Redundancy Limitations

When using ISP redundancy with load-balancing, there are a number of limitations where routing comes in to play, I’ll try to bullet point a few rules:

* In general, traffic responses will be routed back through the pipe the requests went out on. (you can force it to do otherwise but then you are into asymmetrical routing and you will be lucky to get your responses.)

* Anything with a static NAT will be forced out through the 1st link

* You can force certain services to go through your first link but you cannot force anything to go through your 2nd link – see here for instructions.

* You cannot use hide NAT to force anything to go through one link or the other – Checkpoint per se does not support source-based routing (policy-based routing)

* You can in theory force one particular machine to route through the 1st link by statically NATing it to a free public address on your subnet but you will need to have 1 routable IP per machine for which this is to be done. You cannot  force any traffic / services through your 2nd link.

If you really need the flexibility to route your traffic based on services and/or source IPs then you can:
* install a router in fron of your firewall which does policy-based routing

* if you are running on linux or SecurePlatform then you can configure the iproute2 daemon (this will NOT be supported by Checkpoint)

* if you are running on Nokia IPSO boxes then Policy-based routing functionality is built in from IPSO 4.2 build 069 onwards.

Leave a ReplyCancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Exit mobile version
%%footer%%