Tomi Engdahl’s ePanorama blog

IPv6 has been considered the internet of the future for the past 15 years. I played with it in the later 1990’s. And after it waiting when it will take on, or will it fail. Disruption might be the only way to force companies into deploying IP version 6 (IPv6) networks. The expected ending of IPv4 addresses has not caused panic. World IPv6 Day coming at 8 June, 2011 (after a few days) might be the extra disruption that is needed. The goal of the Test Flight Day is to motivate organizations across the industry to prepare their services for IPv6 to ensure a successful transition as IPv4 addresses run out.

Google, Facebook, Yahoo!, Akamai and Limelight Networks will be amongst some of the major organisations that will offer their content over IPv6 for a 24-hour “test flight”. I expect to see some hiccups, although most of the things should work without problems. Google expects less than one per cent of requests to fail on IPv6 day. The vast majority (99.95%) of Internet users shouldn’t see anything out of the ordinary. For the remaining 0.05% of users, misconfigured or misbehaving network equipment, particularly in home networks, may impair access to participating websites during the trial.

You can test your IPv6 connectivity in advance with http://www.kame.net/, http://test-ipv6.com/, Wireshark IPv4 and IPv6 Connectivity Test or ipv6test.google.com. Don’t worry if you don’t have IPv6 at all (just “old” IPv4 connectivity), you shouldn’t have problems on websites that add IPv6 support on World IPv6 Day.

I still like to follow what is happening IEEE is doing on Ethernet standardization, although there has been around 10 years since I last time was part of the process. I was participating in Ethernet in the first mile (EFM) standardization work at years 2000-2001 (at that time I worked for Nokia).

There is always something new coming. Now IEEE Seeks Data On Ethernet Bandwidth Needs. The IEEE has formed a group to assess demand for a faster form of Ethernet, taking the first step toward what could become a Terabit Ethernet standard. This time around, the information gathered may help the next high-speed Ethernet group decide whether to aim for 1T bps or 400G bps. There are much more significant challenges involved in achieving Terabit Ethernet than there are in 400G bps. The ad hoc group has met a few times since late February.

There is also push is on to cut 100G Ethernet’s price. Less than a year after 100-Gigabit Ethernet was standardized, an industry group is considering a set of specifications that might make the high-speed technology less expensive and more useful. The IEEE 802.3 100-Gigabit Backplane and Copper Cable Study Group of the IEEE try to make it easier to build modules with more 100GE ports. The effort to develop specifications for 100GE backplanes and narrow cable interface. Standardization of the backplane should help 100GE move beyond proprietary designs and create a larger ecosystem of component vendors.

Today is a historic milestone for the Internet: the allocation of the last remaining IPv4 Internet addresses from a central pool. The ICANN pool of available unallocated addresses for IPv4 is completely depleted.

The allocation of the final IPv4 addresses is analogous to the last crates of a product leaving a manufacturing warehouse and going to the regional stores or distributions centers, where they can still be distributed to the public. Once they are gone, the supply is exhausted. It’s only a matter of time before the Internet Service Providers (ISPs) must start denying requests for IPv4 address space.

This is not the IPocalypse. The Internet technical community has been planning for IPv4 depletion for some time. IPv6, a new Internet Protocol with a massive amount of address space, is already taking over as IPv4 runs out. For most users, all it means is that your computer’s IP address today might look like 192.0.2.10 (an example IPv4 address), but soon it may resemble 2001:0DB8::/feed:b766 (an example IPv6 address).

The next big event for IPv6 deployment is World IPv6 Day on 8 June, 2011. The goal of this Test Flight Day is to motivate organizations across the industry – Internet service providers, hardware makers, operating system vendors and web companies – to prepare their services for IPv6 to ensure a successful transition as IPv4 addresses run out. Want to find out your IPv6 readiness? Use this test.

Ethernet link speeds of 100 megabits per second or even 1 Gbit/s per second are typical in today’s LANs. When everything is wired to Ethernet LAN very many LAN connections are needed and they consume some power. And the faster the Ethernet connection is, typically more power it consumes.

IEEE Spectrum Energy-Efficient Ethernet article tells that Ethernet connections waste lots of watts and it need not necessarily need be so. On average, people use their Ethernet links at full throttle less than 5 percent of the time. But the circuitry on the network-interface controller, the chip that connects your computer to the network, is always running at full speed, thus wasting power.

One seemingly simple solution is to adapt the Ethernet link’s speed to match a device’s needs. This concept is called Adaptive Link Rate. The problem in this concept is that switching between Ethernet speeds is time-consuming. When you change link rate today, you have to drop the link and reestablish it, which takes up to 2seconds.

Another concept is called low-power idle, it proposes transferring data on an Ethernet link at the highest possible rate and then putting the network controller chip into a sleep-like state. The trouble is that turning on a dormant network card quickly is a challenge, but it is easier than switching between rates.

Wikipedia Energy Efficient Ethernet article tells that Energy Efficient Ethernet, also known as IEEE 802.3az, is a set of enhancements to the twisted-pair and backplane ethernet networking standards that allow for less power consumption during periods of low link use. The goal is to reduce power use by 50% or more, while remaining fully compatible with existing equipment. The power reduction is accomplished in a few ways:

• For 100 Mbit/s and gigabit speed links, ethernet chips that don’t have data to send would put the physical layer of the system into sleep mode. Gigabit interface card might be able to reduce its power by up to 1.5 W.
• For 10 Gbit/s links, speeds are stepped down to slower speeds saving 10 – 20 W per link (twisted pair Gbit/s links consume typically 15W or so power)
• A new lower voltage mode was added to 10Base-T (lower voltage used when full voltage is not needed for example on shorter links)

There were some companies introduced technology to reduce the power required for Ethernet before the standard was set. The best known of them was D-Link’s and Broadcom’s Green Ethernet, which is a superset of the draft 802.3az. Green Ethernet was first employed on home switches and smart switches. D-Link claims that a power savings of up to 45 – 80 percent can be made using its Green Ethernet switches, but I think in real life applications the savings will be somewhat less than those promises.

Broadcom delivers energy-efficient Ethernet silicon article about Broadcom’s portfolio of available silicon supporting the newly ratified IEEE 802.3az-2010 Energy Efficient Ethernet (EEE) standard and proprietary AutoGrEEEn technology. roadcom Infrastructure & Networking Group, said in a press release, “Energy efficiency is of crucial importance to our customers and to the networking industry as a whole.”

The power savings for an individual consumer level device are likely to be modest compared to the cost of device or the embodied energy, so turning off existing devices when they are idle is likely to be a more immediate savings. If Energy Efficient Ethernet is widely incorporated into new systems, it could reduce networking power requirements as systems are replaced. Green Ethernet Technology is pushing into networking hardware market.