- May 29, 2011
Multiple physical network interfaces on the same subnet may cause routing issues
the key word in this document is ROUTING. What is routing? Routing is needed when you want to access another subnet.
No, it is needed when you want access to ANY subnet, including a directly connected one. Routing is the process of selecting an appropriate interface out which to pass a packet, and, potentially, also a next-hop.
RFC1812, 2.2.3, "Forwarding an IP datagram generally requires the router to choose the address and relevant interface of the next-hop router or (for the final hop) the destination host. This choice, called relaying or forwarding depends upon a route database within the router. The route database is also called a routing table or forwarding table."
See that? "or (for the final hop) the destination host". You need the routing table to get to that too. Want to try a neat trick? Configure up an IPv4 interface, then issue a "route delete" for the directly connected network. You will lose access to the directly connected network.
The rest of your message is similarly off-kilter, derived from your basic initial misunderstanding. There's a massive difference between things kinda-working because you haven't done something sufficiently broken for it to actually break disastrously, and something that's being used the way it is designed to be.
This is a PERFECTLY valid network configuration; but there is a challenge.
It is an invalid network configuration. That it works is a testament to how resilient IP networking is and how well-written the BSD IP stack is.
It's not an "issue" or a "problem", but a challenge.
It is not an issue or a problem OR a challenge. It's just broken. Don't do it. It won't work the way the average person might wish. Modern standards allow for the use of link aggregation to accomplish this. LACP works better than this backwards strategy of trying to get multiple physical interfaces configured as logical interfaces on a single network.
Now, please note, the article you linked does include the following:
Now there will invariably be some people who will say ... "WRONG! I know that multiple interfaces are legal!" At which we come to an impasse about what kind-of works versus what actually happens.
I don't actually care to debate with you why you feel multiple non-LACP interfaces "should" work a certain way. I don't actually care if it worked for you under Windows or on some storage platform. I'm here to document the way it works on FreeBSD, Linux, Apple, and most other UNIX variants. I didn't make these design decisions but I am spending some time here to help you understand them. So please don't belabor the point.
I'm dead serious. This IS the way that it is. What you're promoting is an invalid network configuration. No amount of insistence otherwise, or hacky workarounds, can make it magically valid. This not a "challenge" to overcome. Your way is broken. I'm sorry. But I'm also not going to entertain allowing blatantly incorrect information being posted to the forum. We have too many new users who will latch onto the vaguest wisp of a hope that something they want to be able to do will "work," and then we end up needing to keep cleaning up the results of your misinformed contribution, long after you've stopped participating in the forum, every time another newbie comes along and tries what you suggest.