Thank you all for details, not that they were unexpected - but at least these responses confirm that it's not my curly hands at fault ;)
Regarding the notion that "IP doesn't route based on source address", I know that, but somehow expected that source IP is somehow considered when choosing a correct gateway for it. Specifically, if I have two interfaces, for example 192.168.100.10 and 192.168.200.10, and two default routers in these same subnets, say 192.168.100.1 and 192.168.200.1, I did not expect to see packets from 192.168.100.10 trying to go out via 192.168.200.1 router. At the L3 OSI layer these addresses on two sides of the links would not even "see" each other. Alas, for routers L2 often comes into play, so ethernet frames can go out with any random IP addresses - and they do reach the router in the segment, only to be dropped. So, as far as I can tell from the respected gurus' responses, this behavior is expected, works as designed, and won't be fixed. Correct? I'll take a shot at ipfilter reinjection, I guess, or at separating the services into two zones serving the same data with the same software on different networks... -- This message posted from opensolaris.org _______________________________________________ networking-discuss mailing list [email protected]
