Coffeehouse Thread

5 posts

Four Ways IPv6 Will Save the Internet

Back to Forum: Coffeehouse
  • User profile image
    peterwillcn

    The world is almost out of IP addresses--or at least it's almost out of the IPv4 addresses that IT admins and users are most familiar with. Fortunately, IPv6 has been developed to exponentially expand the pool of available IP addresses while also providing a few other benefits.
    To address issues with the current IP protocol in use (IPv4), and to add features to improve the protocol for the future, the Internet Engineering Task Force (IETF) has introduced IPv6 (Internet Protocol version 6). Let's look at a few ways that IPv6 will come to the rescue and save the Internet.

    1. More Addresses. IPv6 uses 128-bit addresses rather than the current 32-bit addresses allowing for an exponential increase in the number of available IP addresses. Network routing experts have been aware of the limitations of IPv4 addressing since the 1980's--before most of the world even knew the Internet existed, and before the Web became ubiquitous.

    With 32-bits, IPv4 only allows for approximately 4.3 billion IP addresses. Since every Web-connected device--server, desktop, notebook, smartphone, tablet PC, and more--must have an IP address, and since the Web has exploded in countries like China and India, the world is rapidly approaching the end of the available IPv4 addresses.

    With 128-bits to use for creating unique addresses, IPv6 is capable of an insane number of IP addresses far exceeding the demands of today's Internet and Web connected devices. Rather than 4.3 billion address limitation of IPv4, IPv6 has enough addresses that every single person on the planet could have billions upon billions assigned to them personally.

    2. Backwards Compatible. IPv6 is backwards compatible with IPv4 so that different networks or hardware manufacturers can choose to upgrade at different times without disrupting the current flow of data on the Internet. Given the world's dependence on the Internet and Web for news, commerce, public safety, national security, and more this is an important feature of IPv6.

    Eventually, all IPv4 equipment will be replaced by attrition and IPv6 will be the only address protocol in use. But, that could take decades, so backwards compatibility will be crucial for the foreseeable future.

    3. Better Security. IPv6 was designed with encryption and authentication in mind. IPsec is an optional security component of IPv4, but in IPv6 it is mandatory. With IPsec each individual data packet is encrypted and authenticated, making many of the malicious attacks plaguing the Internet today impossible--or at least substantially more difficult.

    4. Better Performance. Changes have been made to the way IP packets and headers are formed, and to the way IPv6 routers process the packets to improve performance--resulting in fewer missed or dropped packets, and more reliable and efficient connections. With more people and devices sharing the Internet, and higher demands for VoIP (voice over IP) and video streaming, performance will be more critical than ever.

    It has been coming for more than two decades, but the IPv4 address crunch is beginning to get very real. Expect to see a spike in IPv6 adoption in the near future.

  • User profile image
    Yggdrasil

    1994: RFC submitted by Apple, SGI and Sun bemon the inherent problems in the network infrastructure running out of IP addresses.

    1998: Vint Cerf warns that the internet will run out of IP addresses. Pushes for adoption of IPv6.

    2010: Same Vint Cerf once again warns that we'll run out of addresses within the year. I remain skeptical.

     

    It's not that IPv4 won't run out of addresses eventually. It's just that IPv6 is obviously a bad solution, since vendors have resisted implementing it for over a decade.

  • User profile image
    W3bbo

    Yggdrasil said:

    1994: RFC submitted by Apple, SGI and Sun bemon the inherent problems in the network infrastructure running out of IP addresses.

    1998: Vint Cerf warns that the internet will run out of IP addresses. Pushes for adoption of IPv6.

    2010: Same Vint Cerf once again warns that we'll run out of addresses within the year. I remain skeptical.

     

    It's not that IPv4 won't run out of addresses eventually. It's just that IPv6 is obviously a bad solution, since vendors have resisted implementing it for over a decade.

    Vendors would have resisted implementing anything because it's hard to convince upper management that implementing a feature that no-one can use immediately would be somehow profitable.

     

    We got lucky with x64 because with AMD being the first to implement 64-bit x86 it meant they could set the standard and collect royalties from everyone else (i.e. Intel), but there's no advantage to making IPv6-compatible networking hardware because it will simply cost the company money and not earn anything back.

     

    Here's my realistic scenario for how the Internet's going to shape up the next 5 years or so:

    • Some time in 2011 - All IPv4 blocks are allocated. Some companies with larger blocks will begin auctioning off parts of their address space.
    • 2011-2012: As IPv6 is rolled out (at the last minute) on the backbones, consumer ISPs find themselves unable to allocate addresses to new customers, as a result more of them begin ISP-wide network address translation. Some of the more unscrupulous ISPs charge extra for a public address. This has the effect of boosting ISPs operating revenue and means they can put off IPv6 roll-out for a long while.
    • 2014-2015: IPv6-compatible consumer networking appliances finally become commonplace

    I predict IPv4-over-IPv6 tunnelling will become popular for legacy applications that were hardcoded to use IPv4, but I don't think the backbones will become IPv4-free for a long time.

     

  • User profile image
    intelman

    W3bbo said:
    Yggdrasil said:
    *snip*

    Vendors would have resisted implementing anything because it's hard to convince upper management that implementing a feature that no-one can use immediately would be somehow profitable.

     

    We got lucky with x64 because with AMD being the first to implement 64-bit x86 it meant they could set the standard and collect royalties from everyone else (i.e. Intel), but there's no advantage to making IPv6-compatible networking hardware because it will simply cost the company money and not earn anything back.

     

    Here's my realistic scenario for how the Internet's going to shape up the next 5 years or so:

    • Some time in 2011 - All IPv4 blocks are allocated. Some companies with larger blocks will begin auctioning off parts of their address space.
    • 2011-2012: As IPv6 is rolled out (at the last minute) on the backbones, consumer ISPs find themselves unable to allocate addresses to new customers, as a result more of them begin ISP-wide network address translation. Some of the more unscrupulous ISPs charge extra for a public address. This has the effect of boosting ISPs operating revenue and means they can put off IPv6 roll-out for a long while.
    • 2014-2015: IPv6-compatible consumer networking appliances finally become commonplace

    I predict IPv4-over-IPv6 tunnelling will become popular for legacy applications that were hardcoded to use IPv4, but I don't think the backbones will become IPv4-free for a long time.

     

    What scares me is your scenario sounds plausible.

  • User profile image
    PaoloM

    Considering that you obviously didn't type this out of your brain, it would be nice to at least attribute where you found it.

     

    (it'd be here)

Comments closed

Comments have been closed since this content was published more than 30 days ago, but if you'd like to continue the conversation, please create a new thread in our Forums, or Contact Us and let us know.