Imagine for a moment that there were a nation-wide shortage of phone numbers. The 10-digit numbering scheme just isn't cutting it anymore. The powers that be decide to expand the scheme, but rather than just tacking on a single digit and thus expanding the available number pool by a factor of 10, they decide to think big. REALLY big. Instead of the base 10 decimal number scheme where every digit is 0-9, they go with hexadecimal where every digit is one of 16 characters (0-9 and a-f). And not only so, but they expand the number of digits from 10 to 32.
So now instead of a phone number looking like 408-526-7209, it looks like 2001:a418:1e25:0280:ab3c:f121:00d1:90a2.
IPv6 is kinda like that.
Actually, IPv6 is exactly like that.
What would be the implications of such a shift? Well, I can think of a few:
- Those annoying radio commercials where they repeat the phone number four times would become REALLY annoying.
- Unless you're this guy, remembering phone numbers won't be easy.
- Dialing phone numbers would take a lot longer.
- You'd get a lot more frustrated when you keep fat-fingering the number and have to constantly start over.
Most engineers, when working in one environment for any length of time, become familiar with or accustomed to the IPv4 addressing scheme in use, eventually memorizing dozens if not hundreds of IP addresses for commonly accessed devices. These engineers, when tasked with coming up with an IPv6 addressing scheme for the enterprise, all too often try to approach the task like IPv4. Striving to maintain the familiarity they've gained over the years, they come up with all sorts of crazy methods, like embedding part of the old IPv4 address space into portions of the IPv6 address, or encoding an incompatible number scheme (like Vlan identifiers) into part of the address.
A better way
These misguided attempts at creating a human-readable IPv6 addressing scheme invariably result in a disorganized and sloppy addressing structure. Is there a better way?
Of course there is! In our fictitious phone number scenario, what would be the logical alternative to trying to memorize the new numbering scheme? Why, speed dial of course! You wouldn't try to memorize dozens or hundreds of phone numbers; you'd program them into your phone so you just hit the person's name or picture to call them instead of typing in their phone number.
The solution is the same for networking, except instead of speed dial, it's DNS. Network engineers need to become accustomed to accessing their devices using human-readable DNS names instead of by IP address. By shifting to DNS, engineers can be freed up to design a nice, well-thought-out IPv6 addressing scheme without worrying about how in the world they will remember all those addresses.
Why does it matter? What does it mean for the future of the enterprise?
IPv6 presents a huge opportunity, one that we will never again see in our lifetimes, and one that probably the next 10, indeed likely the next 100, generations won't see. Don't squander the opportunity to create a new, nicely laid-out addressing scheme that can serve your enterprise much better than the haphazard, patched-together IPv4 address scheme you have now. So find your most OCD engineers and have them come up with the best addressing scheme they can without thinking about how people will be able to memorize it.
This article is published as part of the IDG Contributor Network. Want to Join?