Thinking about the iPhone 6

▼ It looks like the introduction of the iPhone 6 is only a few weeks away. And if the rumors are to be believed, it's going to be bigger than the iPhone 5/5S/5C. People talk about 4.7 inches vs 4 inches for the iPhone 5 family, which I'll take as 12 cm vs 10 cm—a 20% increase¹.

That 10 cm for the iPhone 5 is the screen diagonal, which means that the screen is 5 by 8.8 cm. The size of the iPhone 5C is 5.7 by 12.4 cm, so it's 0.7 cm wider and 3.6 cm taller than the screen. If the iPhone 6 also has a screen with a 16-by-9 aspect ratio, that means the screen will be 6 by 10.6 cm. So if nothing else changes, the phone would be 6.7 cm wide and 14.2 cm tall. However, it makes sense that Apple would be able to trim the bezel, especially the "forehead" above the screen—it will be harder to shrink the "chin" now that it houses a fingerprint scanner. So the iPhone 6 is going to be big, but not huge. I currently have a case for my iPhone 5 that's 13.2 by 7.1 cm, and that's entirely workable.

I'm certainly interested in having a bigger phone, as my current iPhone 5 is still very cramped when it comes to reading text and watching video. Some people argue that a big phone is problematic because you can't reach the corners of the screen onehanded. I find that a nonsensical argument. There's no law that says touch targets need to be in the far corners of the screen. In that regard, it's strange that buttons (well, to the degree that iOS still has buttons) tend to be at the top of the screen these days, while earlier they used to be at the bottom, within an easy thumb's reach. However, the now-common left-to-right swipe to get back to the previous screen helps a lot.

But the question is: if the new iPhone has a larger screen, how will the extra space be used? To fit the same amount of text, but in larger type, so it's easier to read, or to fit more text at the same size? There's something to be said for both, so I hope that Apple will make this a user choice. (They already sort of do that by letting you select a text size preference in iOS 7.) A related question is what the pixel dimensions of the iPhone 6 will be. There are basically four choices:

1. 1136 x 640: the same number of pixels as the iPhone 5/5S/5C, 106 pixels/cm (271 pixels/inch)
2. 1364 x 768: the same pixel size as the iPhone 5/5S/5C at 128 pixels/cm (326 pixels/inch)
3. 1280 x 720 HDTV resolution at 120 pixels/cm (305 pixels/inch)
4. 1920 x 1080 HDTV resolution at 180 pixels/cm (457 pixels/inch)

I'm thinking (and hoping) that it's not 1136 x 640. Apple has refused to play the "let's go to insanely high resolutions" game that has been going on on the Android side, but dropping below 300 PPI while some Android phones are pushing 500 would probably be too much of an alleingang, in my opinion. I also doubt that Jony Ive and his team would allow ugly pixels to be visible. The advantage would be that no changes to iOS applications would be required.

1364 x 768 would be an obvious choice, as there's no hit in pixel resolution. Also, 768 is a familiar number from the iPad screen. However, I don't think there's much synergy there as text size and touch targets intended for the iPad would be way too small on an iPhone—the screen of the iPad mini is twice as wide (12 cm) as that of the assumed iPhone 6 (6 cm). Also, how would existing iPhone apps run? Scaling 1136 x 640 to 1364 x 768 would probably not look great, and not scaling means that there's a 0.5 cm border on the sides and 0.9 cm border on the top and bottom of the screen (in portrait orientation). Far from ideal.

I think 1280 x 720 makes the most sense. I'm not sure to what degree Apple values using round numbers, but 1280 x 720 is a widely used HDTV resolution, which can't hurt. At these pixel dimensions, the screen resolution would be still be a hair over 305 PPI, which is the magic "Retina®" resolution. (As far as there is such a thing.) Unupdated iPhone applications could be displayed with borders on all sides, which would be 0.3 cm left and right and 0.6 cm on the top and bottom (portrait mode). That seems fairly workable—the iPhone 5 had almost 0.7 cm borders at the top and bottom when running unupdated apps.

Finally, there's the possibility that Apple decides if a 1280 x 720 HDTV resolution is good, the higher 1920 x 1080 HDTV resolution would be great. I highly doubt they would do this, as such a high resolution provides diminishing returns, but does require a good deal more processing to push more than twice the number of pixels around. Also, the new pixel resolution would be 1.6875 times bigger than the old one, requiring non-integer scaling of applications. Or with small black borders, the scaling factor could be 1.5 times for "retina" apps and 3 x for low resolution applications. That just doesn't seem The Apple Way. Then again, last year Apple decided it was time for 64-bit computing in the mobile space, and nobody saw that coming.

Hopefully, the iPhone 6 comes with 2 GB RAM. The current 1 GB RAM isn't as tight as 128 (!), 256 and 512 MB in various earlier iPhones, but 64-bit code as well as the duplication required to run 32-bit and 64-bit code side by side eat up a good amount of extra RAM and it's good to have some reserve to power future iOSes.

It's high time that 16 GB as the default storage option goes away. On my iPhone 4, 32 GB was pretty spacious, although even four years ago, paying € 100 extra for the upgrade from 16 GB was expensive. But iOS has bloated a lot, and with a few big applications and 16 GB music, I barely have room for a video or two. So I really need a 64 GB iPhone, and I'm not going to pay € 200 extra for that. You can buy a 500 GB SSD for that kind of money these days.

Last but not least, I hope Apple can resist making the iPhone 6 ultra-thin, and leaves room for a decent battery instead. When I got my iPhone 5, I did a bunch of battery tests, and at that point, the battery life was pretty much identical to that of my iPhone 4, which was two years old by then. However, pretty soon I started having issues where the phone would shut down if I tried to do too much at the same time if the battery was below 50%, especially in cold weather. (Taking a photo would typically do it, as that fires up the GPS which in turn causes cellular data traffic.) It took me a while to figure out that five minutes later, I could turn the phone back on. These days, I have to be very careful with the way I use my iPhone in order to make it last through the day on a single charge.

I've been buying new iPhones every two years since 2008, so I could very well buy the iPhone 6 this year. (And it certainly helps that a new Apple Store has opened pretty much within walking distance from my house!) On the other hand, the even years bring unproven redesigns. Two years ago, there was the lightning cable annoyance, and it took Apple a year to get around to releasing new docks. So I could also see myself skipping a year, replace my iPhone 5's battery instead, and see what the iPhone 6S or 6C brings next year.

¹ Update: actually it's 10.2 cm, so going from the iPhone 5 family to a 12 cm iPhone 6 constitutes a 18% increase in size, not 20%. Also see this post.

Permalink - posted 2014-08-21

BGP table hitting 512k limit in older routers

▼ It's never a good sign when the regular press reports about BGP-related issues, such as last week: Browsing speeds may slow as net hardware bug bites (BBC). The problem is that the BGP table has started to hit the 512k FIB limit in some older routers. Numerous outages and slowdowns were reported to be caused by this, but it's unclear to which degree that's accurate.

So what's a FIB and why would it be limited to 512k (524288) prefixes?

BGP routers actually have (at least) three different tables where IP address prefixes are stored, along with a next hop address: the BGP RIB, the main routing table / RIB and the FIB. The BGP Routing Information Base collects all information received over BGP that's not immediately filtered out. So if you have two ISPs, they will both send all the prefixes for all networks in the world that are currently reachable—which means that your router has two copies of every prefix, one with a next hop address pointing to ISP A and one with a next hop pointing to ISP B.

For each prefix, BGP then decides which path is better and sends the prefix plus next hop address pointing to either A or B to the main routing table. The main routing table also holds non-BGP routing information. In large networks, all the internal stuff and routes for customers can add up to thousands or tens of thousands of routing table entries.

Finally, a Forwarding Information Base (FIB) is constructed from the main routing table that is used to actually forward packets to the router identified by the next hop address. Some routers use regular RAM to store the FIB, others use a Ternary Content Addressable Memory. RAM sizes are pretty large these days and typically don't have a fixed limit, as it's shared by many processes running on a router. But TCAMs are special memories with a tiny bit of processing power. Basically, you can show a prefix to a TCAM and then the TCAM will tell you the address where that prefix is stored—you don't have to search through the memory one step at a time. This means TCAMs are very fast, but they are also more expensive than RAM and they run fairly hot. So TCAM sizes are limited.

Nothing new under the sun

Cisco 6500 and 7600 modular routers/switches used to have supervisor modules with a TCAM limit of 256k. And then in 2008 the routing table grew to 256k, so people had to upgrade in a hurry. If they bought new supervisor modules that can handle 512k, they got six years of use out of those, hence Geoff Huston's statement that "Nothing in BGP looks like it's melting".

Because different networks have different numbers of internal prefixes and there are also slight differences between the number of prefixes each ISP announces to its customers, different people get bitten by the issue at different times. Also, TCAMs are often partitioned into different parts: one for the IPv4 routing table, one for the IPv6 routing table, one for MPLS, one for filtering... In some cases, simply changing the partitioning is enough to get by for a while. Alternatively, it's always possible to filter BGP prefixes. As Randy Bush says:

❝half the routing table is deagg crap. filter it.❞

The trouble is, you then lose connectivity towards the filtered prefixes, and there is no obvious way to only filter out the prefixes that are unnecessary deaggregation. If your network is non-huge, the solution is to use a default route pointing to your ISP / one of your ISPs as a safety net. What I used to do many years ago when using severely underpowered routers to run BGP is simply filter out all AS paths longer than five ASes from both our ISPs. Then, if one ISP has a long path and one a short path, I'd still have the short path which I'd want to use anyway. If neither had a short path, chances were it was a non-critical prefix far away, so handling it through a possibly non-optimal default route was unlikely to be problematic.

However, large networks don't have anyone they can point a default route to. So they have to have more recent routers, and they pretty much always do. However, it's not unheard of for older network equipment live out its final years in far away corners of big networks, so they could still have minor issues.

Mea culpa

Although during my training courses, I always warn people that they should buy routers big enough to hold enough prefixes for some years to come, I really should have been more explicit and posted a warning here on this site. At least Cisco did: The Size of the Internet Global Routing Table and Its Potential Side Effects.

The future

Geoff Huston expects the IPv4 table to hit 1 million in 2019 and recommends buying routers that can handle at least 2 million prefixes. Unfortunately, it's not always obvious how many prefixes a router can handle, especially if the TCAM is used for more than just the IPv4 FIB. So make sure what the limits are before you spend your money. Also, keep an eye on the weekly routing table report so you can take action when the BGP table starts creeping up to your routers' limits.

Further reading:

• Global Internet Routing Table Reaches 512k Milestone (Cisco blog)
• What caused today's Internet hiccup (BGPmon)
• Internet Touches Half Million Routes: Outages Possible Next Week (Renesys blog)

Permalink - posted 2014-08-19 - 🇳🇱 Nederlandse versie