I am thinking of upgrading our networking situation and am considering a migration of IPv4 to IPv6. Is it too soon to be considering this move as it looks like much of the networking gear and OSen are ready for it? Where do I lose out in such a situation?
TL;DR: If you're a consumer/SOHO nothing needs doing. If you're providing services, start planning now while there's no rush and no deadline.
Basically, everything is dual stack these days, meaning it supports IPv4 and IPv6 out of the box. The Internet in general is still operating fully supporting IPv4 as much as possible.
At this point, if you're providing services to the Internet -- running DNS, web, email, etc. services. You should be planning to provide those services as both IPv4 and IPv6.
If you're consuming services from the Internet -- small office/home office, larger offices, residential, etc. then there is currently no urgency to convert yourself to IPv6. I suspect as the bigger cable companies start deploying IPv6 the "last mile." Then we'll see more IPv6 traffic in general.
Today, there are a lot of really big things -- Amazon, some of the content delivery networks -- which are only IPv4. So the focus, as I see it, is for the providers to continue working to provide evertyhing in IPv4 and IPv6 simultaneously. The first problem the Internet is going to face is fracturing/islands appearing as the newest areas, where IPv4 addresses are already exhausted, come online as IPv6-only. Those newest areas will need the existing services and service providers to offer stuff over IPv6.
Usually you don't switch from IPv4 to IPv6, you go from IPv4-only to dual stacked networking: IPv4 and IPv6 combined. The IPV6-only internet is very, very limited at this moment.
It certainly isn't too soon to start implementing IPv6: we're rapidly running out of available IPv4 space.
In general, yes. If your systems (routers, workstations, servers, etc) support IPv6 and your ISP provides IPv6, then there's very little reason not to move forward, 'tho you'll want IPv4 to remain your default protocol. And we're talking "dual stack" (v4 and v6), not "turn off IPv4".
There are parts of the world today that simply have no more IPv4 addresses; IPv6 is their only means of connectivity. In order to communicate with them, you'll have to be on the IPv6 internet -- or they need an IPv4-IPv6 proxy, as the protocols are not compatible. (or they're stuck behind massive carrier-grade-nat (CGN) systems.) Unless you're doing a lot of business with Asia, I doubt you've noticed. (or need to)
The only current drawback that I'm aware of is that the IPv6 internet is somewhat less stable and slower. Because many providers aren't focusing as much effort on IPv6, they have fewer IPv6 peering connections over slower alternate links, or worse, tunnels.
[My IPv6 is over a HE.net tunnel. Stock android devices prefer IPv6, and they're HORRIBLY slow to various google properties (play store, youtube) over that tunnel.]
I think you should first opt to run a dual stack, meaning to run IPv4 and IPv6 next to each other. In that way you can steadily progress to an all IPv6 environment once IPv4 is getting faced out. Some countries do not speak IPv4 anymore (like China) and others haven't opted in for IPv6 or only partially.
Yes, time to switch is now. A couple of often overlooked issues:
Many IPv4 networks are often overloaded, congested and have asymmetric paths, so you don't even know which side is to blame; for an end-user, having a dual-stack IPv4/IPv6 means that it's much easier to switch routes if a given path is congested. I use IPv6 all the time exactly for this reason; tunnelled IPv6 from he.net is usually of great help, especially since you know it won't be highly asymmetric (so there'd be no guessing which NOC@ to contact if anything goes wrong).
Another benefit of IPv6, especially if you are a content-provider, is that it's much easier to do accounting and tracking, whereas with IPv4 a whole set of NAT'ed customers might all look alike. Many providers of IPv6 even publish detailed info about their /64 and /48 customers in the public whois servers, whereas I've never even heard of a /32 IPv4 SWIP, even a /30 IPv4 SWIP is not something that's generally supported by anyone.
The question is now two years old, but the issue still exists.
My answer assumes that you are on the server side, although similar reasoning also applies on the client side.
You would do well to get ready for IPv6. Make sure your you have IPv6 allocations. Test all your systems. Get to the point where you can simply flip the switch by adding AAAA records to your DNS.
Then stop there and consult with management. The main problem is that the number of support calls will go up dramatically. You will have to deal with two main problems: people whose IPv6 is simply misconfigured, and ISPs whose IPv6 is flaky. As long as most users use IPv4, problems on the IPv6 side may go unreported and undetected for hours.