IETF's Tips For Network Admins On How To Avoid Draining Smartphone Batteries (softpedia.com) 65
An anonymous reader writes: Two engineers from Cisco and Google have raised the problem of IPv6 networks that drain smartphone battery life and issued a series of tips for other network admins on why and how to properly configure their networks. The problem is because of Router Advertisements (RAs). These are periodic messages sent by the router to all network clients telling them its IPv6 address, at which it can be reached. Apparently some networks are sending these as often as every 3 seconds, while the engineers say the proper interval should be 7 per hour. Hence the reason why your battery life is often drained even if your phone is in sleep mode, but connected to a local network.
IPv6 is such a disaster (Score:1)
Have you heard of anything, other than more addresses so you can be sloppy with allocations, positive about it? I haven't.
Re: (Score:3)
Re: (Score:1)
RA is only chatty if you mis-configure it. You can also set the IPv4 DHCP timeout to 5 seconds and complain about how chatty DHCP is.
Why are NETWORK ADMINS responsible? (Score:2)
Precisely! But where in the protocol can one set the frequency of RAs, so that battery life ain't needlessly drained?
But here's the thing. Even if a router advertizes every few seconds, why does a DEVICE have to keep scanning? It just needs to scan when it loses the signal to a particular router, and needs a new IP address. So if I am w/ my iPhone at home, the device shouldn't have to scan at all. If I'm at an airport going from check-in to the terminal where my flight will take off, the device woul
Re: (Score:2)
There are 2 things here - the routers and the devices.
For the routers, since in IPv6, addresses are proactively distributed, as opposed to requested by the device (as was the case in IPv4), RAs SHOULD be frequent, particularly in a wireless environment. Like, say, at an airport, if you have an AP at one of the terminals, where people keep coming and going, RAs should happen frequently, providing addresses to the passengers in the vicinity. If they were to happen rarely, then the passengers who want t
Re: (Score:2)
Looks like they got rid of the idea of a hierarchical routing scheme, which could have helped in this, due to the desire of having Provide Independent IPs
As for IPv6, aside from the addresses, there are a lot of things it brings - like eliminating the necessity of having NAT in cases where having it is a disruptor - like where end to end connectivity is needed. Now, there were a lot of howls about how NAT is good and actually needed in some cases, such as load balancing, and for that, IETF did insert Pr
Re: (Score:1)
Yeah, it makes handoffs between APs a lot easier. Not a big deal, unless you've got clients travelling around the world, connecting through different services at different times in different ways. Works beautifully for smart phones and business travelers.
Re:IPv6 is such a disaster (Score:4, Informative)
An IPv6 network is much easier to set up properly. Check out the HomeNet stuff, where you just chuck a bunch of routers together more or less randomly, with connectivity from cable and DSL and 4G, plus a bunch of wifi routers, and it all Just Works. IPv4 in the same scenario will require a lot of hand-fiddling and being strict about topology.
No worrying about subnetting, setting up DHCP, making sure that there's precisely one DHCP server per network and all that.
Some of it is still under development, like the daisy-chained routers more than one deep, but it will get there.
Re: (Score:2)
Except that no enterprise network engineer in their right mind will use the stateless autoconfig you refer to, because it cannot be secured.
One joker sending his own RAs and you're whole network is p0wned, no spoof protection beyond what your vendor implements in their equipment,
which, just like IPV4, has to be configured. Not that IPv6 is a bad thing, and enterprise gear is now shipping that has the same
level of first-hop security features we've come to expect from IPv4 feature sets. Someday maybe there
Re: (Score:2)
But Access Points are typically plugged in, and are not the devices sensitive to battery life. Those are the iPhones, iPads, Galaxies, and so on. THOSE devices ought to be configured (by default) to only scan for new RAs IF AND WHEN they lose connection w/ their original gateway.
Looks like the IETF is looking at the wrong place to conserve battery life
Re: (Score:2)
Re: (Score:2)
first job of a wifi engineer that cares about not draining batteries is to start to turn off AP propagation of all broadcast and multicast traffic.
Unlike IPv4, Ipv6 doesn't work at all if multicast is filtered -- things like stateless autoconfiguration, DHCPv6 and even neighbour discovery run over multicast.
Re: (Score:2)
First, stateless configuration you just kill off with extreme prejudice.
Second, broadcast is indeed essential for almost every protocol that runs over ethernet (because ARP)
but that doesn't stop us from turning off everything but the bare essentials. Almost all enterprise gear knows
how to let ARP (or ND) and DHCP through while blocking everything else. The better gear even allows you to let
users hear the broadcast from other devices which that user owns, but does not forward them to any other
clients (and
Re: (Score:2)
First, stateless configuration you just kill off with extreme prejudice.
That's up to you, of course, but you still need RAs -- DHCPv6 doesn't distribute a default route, it relies on RAs for that.
Second, broadcast is indeed essential for almost every protocol that runs over ethernet (because ARP) but that doesn't stop us from turning off everything but the bare essentials. Almost all enterprise gear knows how to let ARP (or ND) and DHCP through while blocking everything else.
So you kill DNS-SD over mDNS (Apple's Bonjour)? No printer discovery, no discovery of streaming media servers, no IPTV?
you can also convert all broadcast which you DO allow back out of APs to unicast on the RF side
That's actually a good idea, since multicast over WiFi is horribly inefficient.
Re: (Score:2)
So you kill DNS-SD over mDNS (Apple's Bonjour)? No printer discovery, no discovery of streaming media servers, no IPTV?
Yeah pretty much stomp the crap out of any protocol that is built on the erroneous premise that network segmentation has any bearing on the borders of a LAN/PAN anymore, and then use vendor solutions to filter it and distribute it to specific hosts, both in and outside the IP broadcast domain (the latter part can only be done with protocols that don't have a ban on using servers outside their subnet, so no MS media players.)
Though, demand has not been high enough for my boss to elevate setting that up on my
Re: (Score:2)
An IPv6 network is much easier to set up properly. Check out the HomeNet stuff, where you just chuck a bunch of routers together more or less randomly, with connectivity from cable and DSL and 4G, plus a bunch of wifi routers, and it all Just Works.
The IETF Homenet Working Group [ietf.org]
Non-official Homenet HOWTO [univ-paris-diderot.fr]
Re: (Score:3, Interesting)
The Bad:
-They did away with private addressing (site-local) "because it breaks the openness of the internet and firewalls". Tell that one to someone who's seen hackers use a Java-based PS2 Video broadcasting software to send files across the internet. Lets automatically use public addresses on air-gapped networks.
-The standard has changed so many times in the last 10 years nobody can comprehend it; every book has a different set of material on it, every programmer has set their infrastructure up different
Re: (Score:2)
I'm still waiting for verizon wireless to start giving out routable ipv6 addresses. The At&t dsl line at work has had a valid ipv6 address for the last 3 years.
On a verizon 6620l I am assigned a ipv6 address that is unreachable.
Verizon tops AT&T in IPv6 (Score:2)
Re: (Score:2)
Just to be sure we are talking about the same thing if you for example setup a ftp server on port 21 bound to your ipv6 address can you reach it from another connection with a ipv6 address? I can reach the internet with my ipv6 address it's just the internet doesn't seem to be able to reach me.
Re: (Score:2)
Yes, ISPs try to artificially limit what you can do by making you jump through hoops.
Re: (Score:2)
I am still waiting for AT&T U-Verse to stop blocking IPv6 tunneling over IPv4. They did not used to but when they started selling "upgrades" for IPv6 support, they also started blocking native tunneling.
Re: (Score:2)
Upgrades? The modem at work has native ipv6 its not an added feature.
But just to note it is a business line $45/mo 12/1mbps It may have different features than a residential line. We also own the modem which is unheard of on a residential line.
Hmm I haven't checked if I can still do tunnels over ipv4 I don't see why I wouldn't be able to.
Did they give you a nat ipv4 address instead of a public ipv4?
Re: (Score:2)
Right about they time AT&T started blocking it, they also started advertising IPv6 availability if you payed to upgrade your modem.
This is on a consumer AT&T U-Verse line.
Re:IPv6 is such a disaster (Score:4, Informative)
-They did away with private addressing (site-local) "because it breaks the openness of the internet and firewalls". Tell that one to someone who's seen hackers use a Java-based PS2 Video broadcasting software to send files across the internet. Lets automatically use public addresses on air-gapped networks.
No they didn't.
See https://tools.ietf.org/html/rf... [ietf.org] - Site-local was the original spec and that's deprecated since it doesn't allow for easily merging of two existing private networks. ULA fixes that. So damn right you can have private networks.
The standard has changed so many times in the last 10 years nobody can comprehend it; every book has a different set of material on it, every programmer has set their infrastructure up differently.
Oh please. Only things that have really fundamentally changed are the IPv4IPv6 transition mechanism. Now that NAT64 and DNS64 are in use, you can pretty much work with an IPv6-only network (ironically, a couple years ago everything else, including gaming, worked via a NAT64, except for Skype, which is supposed to go through anything)
They did away with IPV4's simplistic subnetting and supernetting, and introduced EUI-64 addressing which can track devices as they move from network to network. Marketing companies like Google and Microsoft were helping to write the standard.
Oh please, even Windows uses privacy extensions for IPv6. No one forces you to use EUI-64.
Very Few large deployments.
Tell that to the Chinese. They have *huge* networks, IPv6 only.
This. IPv6 unique local = better private addresses (Score:3)
Exactly. Unique local addresses in IPv6 are not globally routable, just like private addresses in IPv4. They can be used just like IPv4 private addresses, if you want to. Unique local is all addressed starting with FD:.
The IPv6 version is better, though, because each local network is likely to use different IPs, so you can choose to locally route them. Here are a couple of examples of why that's good. Suppose you have a small office in College Station, where someone set up a typical SOHO network with uni
Re: (Score:2)
Re: (Score:2)
RFC 1918 addresses were a kludgy solution to problematic address assignment early on, and a recognition that the IPv4 address space was too finite for even the conservative growth forecasts of the time. It
The only place I've seen EUI64 used... (Score:2)
Re: (Score:2)
It was common early on until people (rightfully) started getting concerned about the possibility of device tracking, and it's probably still used in some devices. By the time it started getting implemented at any real scale, privacy extensions became the norm.
The concept has been extended to other areas, too. Apple added it to iOS MAC addressing for hotspot detection (though the real MAC address is used for the actual connection), and the concept has been proposed for IEEE adoption in the 802.11 spec exce
Re: (Score:2)
The Bad: -They did away with IPV4's simplistic subnetting and supernetting
"Simplistic" means "excessively simple", which hardly seems to describe subnetting in IPv4 - should you go for a /27, or a /26, or a /25? IPv6 is simpler, as you don't have to worry about running out of LAN addresses.
-Very Few large deployments.
Comcast had 1Tb/sec 18 months ago [slashdot.org]
Re: (Score:2)
Re: (Score:2)
They did away with private addressing (site-local) "because it breaks the openness of the internet and firewalls"
They did away with site-local addresses because they couldn't agree on the definition of a site (is your home network a site, or is it part of your ISP's network?). They've been replaced with ULAs, which are easier to understand, simpler to administer and simpler to program with.
The standard has changed so many times in the last 10 years
A few unused features have been removed (v4-compatible addressing, site-local addressing, partial support for MTUs below 1280), but the specification has been mostly stable for a good 15 years.
They did away with IPV4's simplistic subnetting and supernetting,
No, subnetting is still there. The on
Re: (Score:2)
Absolutely. It was shockingly easy to setup on my home network. Once I configured OpenWRT with an IPv6 tunnel, and initialized ip6tables, autoconfiguration kicked in, all the devices on my home network, including mobile devices, configured IPv6 on their own and everything just worked.
The article seems to be about misconfigured routers, unrelated to any problems with IPv6 itself.
I'd like to start running IPv6 everywhere. My darned ISP isn't provide it natively, yet, and we don't have it on our work networ
IPv4 can't do it anymore.... (Score:2)
Re: (Score:2)
Re: Short DHCP lease (Score:2)
I remember this on my Bellsouth/AT&T DSL line years ago. It made my system log mostly dhcp renewals. They never did fix it either; dealt with it until I moved to cable.
Comcast may suck in other ways but at least their lease times are several days.
Re: (Score:2)
At&t Arris nvg510 has a dhcp time of 24 hours, 10 minutes if you enable ip passthrough. ADSL2+
Maybe someday our local cable company will be brave enough to admit they don't have speeds as good as dsl on the flyers they send out.
At&t is selling 12mbps while suddenlink is selling 8mbps.
Re: (Score:2)
Unless there's more than one suddenlink their site [suddenlink.com] says they offer 50/5, 100/10, 200/20 and 1000/50 tiers (the gigabit is going to be problematic IMHO as many protocols are going to need more than 5% return bandwidth to work well).
Re: Short DHCP lease (Score:2)
stock firmware on consumer routers has been lowest-bidder crap since the product space began. The *WRT projects, some of Apple's gear, and probably Google's new device would be the exceptions.
I found a 54G the other day at a client site that had be running a DD build from 2009 without missing a beat since it was installed. Makes you feel bad for the hardware engineers who work on these things. We might assume the Belkin/Buffalo/ASUS folks who went open source had hardware engineers who finally ousted th
Hmm... home routers? (Score:2)
I wonder how bad this is on home routers. I haven't seen this problem on my iPhone or Droid but my iPad, only a year old, has recently started draining its battery pretty quickly which I think I can relate to turning on IPv6 at home...
(Or maybe it's just a bad iDevice... :) )
Re: (Score:2)
Oh yeah.. laugh now... just wait til your implant gives you diabeetus because it continuously requires more glucose to keep refreshing its IPv6 connection!
(Besides, old people need the bigger screens, I'm surprised there's not a jitterbug iPhone yet)
Re: (Score:2)
Re: (Score:2)
Since the NSA told all phone manufacturers to keep track of users at all times.
Re: (Score:2)
Yeah, why would someone's communication device need to be reachable on a network? It's puzzling.
Meanwhile at Samsung (Score:2)
Newer Samsung phones are DROPPING any IPv6 packet (not just RAs) as soon as the screen is off. (this is in the WiFi firmware so even 3rd party roms like CyanogenMod are affected).
So anything that is connected via IPv6 will be disconnected.
See this thread (the title is about ICMPv6 but later it clarifies that newer phones drop ALL IPv6 traffic): http://developer.samsung.com/forum/board/thread/view.do?boardName=General&messageId=239890 [samsung.com]
Even if they do that to save battery, there are better ways to d
Router Advertisements. (Score:1)
Re: Router Advertisements. (Score:2)
Some networks have been intentionally configured to send out RAs every 3 seconds because Samsung devices drop all IPv6 when the screen is off. This causes them to lose all network access (even IPv4) when the screen comes back on, until the next RA. Samsung broke it, they need to fix it.
It would of course be better if solicited RAs were not sent as broadcasts.
Re: (Score:2)
How to get more battery life (Score:2)
Okay so if everybody fixes this problem everywhere we're going to get 171.42 days of battery life instead of only one?