iPad fails networking 101; how to earn it a passing grade

The iPad is not playing nice with the DHCP protocol: if the screen is turned off, it doesn't release its IP address when it's supposed to. This leads to address conflicts when the DHCP server gives the address to another system, but there are ways around the problem.

<a href='http://meincmagazine.com/apple/news/2010/04/ipad-fails-networking-101-how-to-earn-it-a-passing-grade.ars'>Read the whole story</a>

 

[David Bradbury]

Seems like a pretty easy fix for Apple -- though this bug shouldn't have made it into the final product.

 

[backtomac]

The guy responsible for fixing this is busy trying to find a lost iPhone.

Cut him a break. He's a little stressed right now.

 

[amardeus]

I wonder how something like this made it to the final product.

 

[jimmyeatapple]

Setting the DHCP lease time on the server to something very high (think 24 hours) will reduce the impact of the problem.

That high of a lease time would be far to much for the university I worked at. We barely had enough addresses to go around, and each year it got worse (and they allotted more space each year, they even considered NATing). It most mostly due to the laptop initiative the college of engineering had, which was timed when I started there. Each year thousands more devices showed up. After four years, there was 6000+ laptops on campus plus untold amounts of smartphones... Sadly, I don't know the specifics on how many were trying to connect at once.

These kinds of bugs kept me employed. I worked at the laptop help desk there.

 

[smellykaka]

I wonder how something like this made it to the final product.

Probably the same way many surprising defects in Apple products show up: extreme secrecy during development means the product is used in a very limited set of circumstances.

 

[ColinABQ]

With the exception of turning WiFi off and back on, I don't think the other listed workarounds are really going to work on large and busy networks, or on networks which the iPad user doesn't own and control. And that's where the solutions are needed most, as home users may not even encounter the problem. Modifying the DHCP server settings isn't an option for most users, nor is using a static IP address without involving the network admins. (You can't just pick one and go for it - you need to know what range is acceptable - and what if the one you pick is already in use? You really need an admin to hand them out.)

Has apple announced a time frame for a patch?

Disclaimer: I'm not an iAnything user, I'm just curious/interested/meddling.

 

[jimmyeatapple]

I wonder how something like this made it to the final product.

This happens all the time with university networks. I cannot tell you the amount of times wireless this and wireless that broke. It was a constant battle to keep printing functioning and often my Powerbook G4 and later MacBook Air would refuse an address. It was also hit and miss with many consumer routers, so I imagine they only have airports at Cupertino

This dude wrote something about wireless printing, which was a lot of fun:
http://www.itunprofessional.com/post.php?post=12

 

[stevenkan]

Like virtually all Internet-capable devices, iPads obtain an IP address using the Dynamic Host Configuration Protocol (DHCP) when they connect to a WiFi network.

Virtually all? Really? Do we have any statistics on this? Because in my house the majority of devices have fixed IP addresses, even on WiFi.

I might believe that statement if it were qualified with "consumer-grade devices" instead of the blanket "virtually all."

 

[rabid77]

This is a pretty stupid mistake on Apple's part. It shouldn't be that hard to code a RFC-compliant DHCP client - they've already done it with the iPhone and their laptops. They shouldn't have introduced this error on the iPad and really shouldn't have missed it in QA.

 

[Pont]

Like virtually all Internet-capable devices, iPads obtain an IP address using the Dynamic Host Configuration Protocol (DHCP) when they connect to a WiFi network.

Virtually all? Really? Do we have any statistics on this? Because in my house the majority of devices have fixed IP addresses, even on WiFi. I might believe that statement if it were qualified with "consumer-grade devices" instead of the blanket "virtually all."

By default, "virtually all" is correct. At least for computers and devices made in the last, say, 5 years.

 

[NicciAdonai]

Like virtually all Internet-capable devices, iPads obtain an IP address using the Dynamic Host Configuration Protocol (DHCP) when they connect to a WiFi network.

Virtually all? Really? Do we have any statistics on this? Because in my house the majority of devices have fixed IP addresses, even on WiFi.

I might believe that statement if it were qualified with "consumer-grade devices" instead of the blanket "virtually all."

Devices in your house aren't capable of DHCP?

 

[NicciAdonai]

@jimmyeatapple:

Cool name.

 

[Salvasian]

...it just works?

 

[Lord Kestrel]

Like virtually all Internet-capable devices, iPads obtain an IP address using the Dynamic Host Configuration Protocol (DHCP) when they connect to a WiFi network.

Virtually all? Really? Do we have any statistics on this? Because in my house the majority of devices have fixed IP addresses, even on WiFi.

I might believe that statement if it were qualified with "consumer-grade devices" instead of the blanket "virtually all."

Every default for every semi-modern device/OS is to use DHCP. While some people change to static, or static dhcp, the majority do not. Once you get past a very small number of devices, managing static IPs is too much of a hassle. From what I've seen in the various networks I've had access to, servers/firewalls/routers are static, everything else is dhcp. And seeing as how the majority of the devices on any given network are likely to be clients rather than servers, it's easy to see where the earlier blanket statement came from.

 

[ZenLabs]

Let me get this straight, your expecting to change the SERVER settings for a fault on one of its CLIENTS? Get real.

 

[affinity7]

I can actually see this occuring easily and getting thru QA....Since the IP stack is alive DHCP service is turned off..Since it wakes up long after the lease is expired it simply goes on probably until the next lease expire or just plain ignores it.
they just need to keep the DHCP client service along with the IP stack, which should be an easy fix....however how long will Apple wait to release the patch.

 

[r3loaded]

Let me get this straight, your expecting to change the SERVER settings for a fault on one of its CLIENTS? Get real.

But..but..it's magical and revolutionary! You also have to set the server up so it's compatible with magical devices!!

 

[siliconaddict]

I wonder how something like this made it to the final product.

Apple has to be first. Enough said.

 

[Joe Buck]

Most bugs look stupid when you first find them, though it's much better for you if you can find them in-house.

The real test for Apple is how quickly they can roll out fixes for issues like this.

 

[MatthiasF]

Can't someone just write a script to restart DHCP or tell it to ask for a new IP when the iPad comes out of sleep mode or if the Wake/Sleep button is pressed?

 

[siliconaddict]

Can't someone just write a script to restart DHCP or tell it to ask for a new IP when the iPad comes out of sleep mode or if the Wake/Sleep button is pressed?

Yah and try and get that passed through Apple's approval system, the patch will be out before its approved. No really.

 

[Enki42]

Can't someone just write a script to restart DHCP or tell it to ask for a new IP when the iPad comes out of sleep mode or if the Wake/Sleep button is pressed?

I'm guessing it would be rejected from the app store

 

[zig_pwnd]

Can't someone just write a script to restart DHCP or tell it to ask for a new IP when the iPad comes out of sleep mode or if the Wake/Sleep button is pressed?

I think that breaks their new license agreement Okay I'm just being a smarta**, ya this shouldn't be a difficult fix. I can't see the renewal breaking anything else but who knows...

 

[hutch85]

What, there isn't an app for that?? (to fix it..?) ;-)

 

[stevenkan]

Like virtually all Internet-capable devices, iPads obtain an IP address using the Dynamic Host Configuration Protocol (DHCP) when they connect to a WiFi network.

Virtually all? Really? Do we have any statistics on this? Because in my house the majority of devices have fixed IP addresses, even on WiFi.

I might believe that statement if it were qualified with "consumer-grade devices" instead of the blanket "virtually all."

Devices in your house aren't capable of DHCP?

Of course they are. I just don't have them configured that way. The article didn't say that virtually all are capable of DHCP; it says that virtually all are "using" DHCP. That's just too general of a statement in a technical journal aimed at a technical audience. Anyway, it's a nitpick, and I'll stop complaining.

 

[DrPizza]

The article didn't say that virtually all are capable of DHCP; it says that virtually all are "using" DHCP. That's just too general of a statement in a technical journal aimed at a technical audience.

The hell it is.

 

[flyingember]

Setting the DHCP lease time on the server to something very high (think 24 hours) will reduce the impact of the problem.

That high of a lease time would be far to much for the university I worked at. We barely had enough addresses to go around, and each year it got worse (and they allotted more space each year, they even considered NATing). It most mostly due to the laptop initiative the college of engineering had, which was timed when I started there. Each year thousands more devices showed up. After four years, there was 6000+ laptops on campus plus untold amounts of smartphones... Sadly, I don't know the specifics on how many were trying to connect at once.

These kinds of bugs kept me employed. I worked at the laptop help desk there.

I'm amazed they weren't using NAT. Must have been a small college. When I was in college they used NAT everywhere.

 

[resistor.one]

Network of 200 machines.

So far 5 iPads.

Have not seen this problem. Suspect reason is deeper than just, "Apple fix it."

 

[zig_pwnd]

41 iPads have been seen and 22 have exhibited the problem, eight to the degree of having been blocked from further access to the network.

I found this a bit confusing. I'm assuming that only 22 of the 41 seen on the network exhibited the problem due to different usage patterns but why were 8 blocked? I can't imagine it's due to deliberate manipulation of this flaw in the iPad on the part of the student, if that was the case they'd have to ban his entire account because he could do the same thing with a custom linux build on a laptop. Not only that but he could just spoof his MAC address and then the device would no longer be "banned."

So were the select 8 banned due to usage patterns, was it due to apps that were installed, was it due to those few being used in certain extremely congested parts of campus or what? Just curious if anybody knows.

 

[dobrien75]

virtually all

means "in general all". i.e. "almost all" or "All except a few cases."

Anyway, it's a nitpick, and I'll stop complaining.

No it's not. It's wrong.

 

[DaemonBlood]

Of course they are. I just don't have them configured that way. The article didn't say that virtually all are capable of DHCP; it says that virtually all are "using" DHCP. That's just too general of a statement in a technical journal aimed at a technical audience. Anyway, it's a nitpick, and I'll stop complaining.

Almost everyone I know uses DHCP. Any devices that require static ips are set using static reservations using their MAC address, which technically still uses DHCP (and still has lease times). I have every device in my house configured this way, and 3000 servers at work but it's still DHCP.

 

[HiRez]

I wonder how something like this made it to the final product.

Probably the same way many surprising defects in Apple products show up: extreme secrecy during development means the product is used in a very limited set of circumstances.

Yep. That's the price you pay for secrecy. It's not really that unexpected, pretty much all version 1 Apple products have some bugs. I remember my original iPhone had some flakiness when it came to networking at first also, hopefully these can be fixed via a software update in the same way. I'm pretty confident they'll get it sorted out.

 

[dobrien75]

It's not really that unexpected, <ALL> version 1 <INSERT SOFTWARE DESIGNER HERE> products have some bugs.

FTFY

 

[daemonios]

Seems like a pretty easy fix for Apple -- though this bug shouldn't have made it into the final product.

It's not a bug. It's a feature. No, really - look how much battery it saves

 

[davis]

Like virtually all Internet-capable devices, iPads obtain an IP address using the Dynamic Host Configuration Protocol (DHCP) when they connect to a WiFi network.

Virtually all? Really? Do we have any statistics on this? Because in my house the majority of devices have fixed IP addresses, even on WiFi.

I might believe that statement if it were qualified with "consumer-grade devices" instead of the blanket "virtually all."

Stop being a dork, obviously they are referring to consumer products. It's a given when they're talking about an iPad. They wouldn't say, relate it to a Ford F150, anymore than they would a Juniper switch.

 

[TheFerenc]

dobrien, you're being generous again. All versions of all consumer products have bugs. Period. Full stop.

Things that are custom produced, never sold, specific intent, whatever can sometimes avoid this, but it's rare. The only example I know of in the software industry is the control software for the space shuttle. Though I don't know how true that is, but the claim was 0 bugs in 500K lines of code.[1]

Interestingly, according to many sources, MS products ship with the one of the lowest values of bugs per 1000 lines of code. But then they ship some of the largest pieces of software, so there is that...[1]


[1] "Code Complete" by Steve McConnell

 

[gameoven]

This remind anyone else of the old comma problem iPhones used to have? When I would run my iphone on wifi, it'd pull down two dns servers through dhcp and separate them with a comma in the form xx.xx.xx.xxx, xx.xx.xx.xxx. It, of course, wouldn't load web pages because it was actually reading the entire string as one dns address, so I'd have to delete the second or the first. I don't get how Apple doesn't notice this stuff in house, it just seems so silly.

 

[davis]

Network of 200 machines.

So far 5 iPads.

Have not seen this problem. Suspect reason is deeper than just, "Apple fix it."

How big is the pool of IPs available to the DHCP server, what's the lease time, and how often do devices "leave" the network?

After those questions, try putting an iPad to sleeping and having the DHCP server assign it to another MAC address and see if said problem appears?

I suspect the reason is deeper as well, but in this case it has nothing to do with the iPad.

 

[davis]

Setting the DHCP lease time on the server to something very high (think 24 hours) will reduce the impact of the problem.

That high of a lease time would be far to much for the university I worked at. We barely had enough addresses to go around, and each year it got worse (and they allotted more space each year, they even considered NATing). It most mostly due to the laptop initiative the college of engineering had, which was timed when I started there. Each year thousands more devices showed up. After four years, there was 6000+ laptops on campus plus untold amounts of smartphones... Sadly, I don't know the specifics on how many were trying to connect at once.

These kinds of bugs kept me employed. I worked at the laptop help desk there.

I'm amazed they weren't using NAT. Must have been a small college. When I was in college they used NAT everywhere.

There is no reason not to use NAT these days, or even better, IPv6. Does the iPad even support IPv6?

Edit: I know there are lots of reasons not to use IPv6 -- sentence didn't come out the way I wanted.