R7800 and issues with iOS/Apple devices on 5GHz

[garethr]

After a more in-depth review of the release notes and security fixes in 10.11.4/IOS 9.3, no changes to the WiFi drivers - some kernel input cleanup to resolve a couple of security issues, but that's about it.

Yeah, I'd not seen anything either which is why I suspected it may be incidental at this point. I guess time will tell over the next few days whether the issue raises its head again.

 

[Killhippie]

Yeah, I'd not seen anything either which is why I suspected it may be incidental at this point. I guess time will tell over the next few days whether the issue raises its head again.

My iPad Air 2 is still having the issue, the iPhone 6s as well, so for myself the update did nothing in respect of connectivity, both devices still connect but a hard reboot caused no connection again, although as before if you put the iPhone or iPad to sleep and then wake it up again its connects. I tried three R7800's with the same issue on every one and that included just unboxing one and attaching it to the wifi with default SSID and no update to the firmware. At this point Netgear are sending the information back to Engineering, to see what's going on. Considering the idevices connect first time perfectly, its probably not something a lot of people have noticed tbh.

 

[JCF]

I've seen a few posts around where people have reported similar issues with Apple devices and the R7800, but no response from Netgear regarding the issue, and whether anything is being done to address it. Right now the R7800 is virtually useless with seemingly any recent Apple devices.

My Apple clients include an iPad Pro, 2013 Macbook Pro, and a couple of iPhone 6S. I have no issues with my other devices.

Essentially when woken from sleep or if wifi is switched off and on quickly on any of the Apple devices, they either refuse to connect to the 5GHz band or take a minute or two of trying to connect before they succeed.

The other (minor) issue is the attached devices reporting is plain wonky on the R7800 - a number of devices appear to fail to report correctly either missing hostnames or even in some cases IP addresses.

I'm on the latest 1.0.0.40 firmware (UK model).

I happen to have two R7800 and this behaviour is consistent between them. I've also used a (Broadcom based) Asus RT-88U which has none of these issues, neither did the Netgear R8000 before it.

I fully except there may be some issues with Apple and there wireless implementation, however given the other Broadcom based products have no issues connecting I'm wondering whether this is a QCA issue, or unique to the R7800.

Has anyone experience similar issues with the R7500v2 for example (which is basically the same hardware)?

I have an issue with my Netgear D7000.
Since I updated my iPhone 6 to OIS 9.2.1, when my iPhone is on wifi, it knocks everything off the wifi and I have no internet access. When I turn of the iPhone 6 wifi, everything works again.
Tried my fixes without success.
I didn't have any problems with my iPhone on OIS 8.2.
Something about the OIS update interferes with my router.

 

[RMinNJ]

Does the log say anything about whats coming from the iPhone devices.

I just got an R7000 I see the log shows a lot of UPnP stuff coming from my android phones. iPhone 5 seems okay so far. I'm worried now about when my kid comes home from college with his ipad. the ipad had network problems on our old 600N linksys sometimes also.

 

[garethr]

Does the log say anything about whats coming from the iPhone devices.

I just got an R7000 I see the log shows a lot of UPnP stuff coming from my android phones. iPhone 5 seems okay so far. I'm worried now about when my kid comes home from college with his ipad. the ipad had network problems on our old 600N linksys sometimes also.

The only thing I see which is slightly odd are multiple clustered log entries relating to DHCP allocation for the affected devices around the times when a device is struggling to connect - whether DHCP is the route cause or not I'm not sure - you'd expect to see multiple DHCP entries each time a device tries and retries to connect. The standard Netgear logs don't really give much in the way of detail as to what is really happening wireless wise.

 

[Killhippie]

The only thing I see which is slightly odd are multiple clustered log entries relating to DHCP allocation for the affected devices around the times when a device is struggling to connect - whether DHCP is the route cause or not I'm not sure - you'd expect to see multiple DHCP entries each time a device tries and retries to connect. The standard Netgear logs don't really give much in the way of detail as to what is really happening wireless wise.

I noticed that, whereas say my PS4 will just show one DHCP request the Apple devices show multiple attempts. Netgear have said their engineers are still looking into this. So time will tell, but it is a very young router.

 

[sfx2000]

Interesting -- if this helps...

Apple's iOS does MAC randomization for 802.11 Probe Request Messages when searching for a known SSID... might have to do a wireshark capture to see this - this is to protect user privacy for situations where AP's are set up for presence detection and tracking (think shopping malls and foot traffic analysis).

If the SSID is a known SSID, it will use the proper MAC address.

 

[NETGEAR Guy]

If anyone on the post would like to test a trial code to address any outstanding specific iPhone issues (inc when using HT160), please let me know.

 

[garethr]

If anyone on the post would like to test a trial code to address any outstanding specific iPhone issues (inc when using HT160), please let me know.

I'd be interested in this. Thanks.

 

[Killhippie]

If anyone on the post would like to test a trial code to address any outstanding specific iPhone issues (inc when using HT160), please let me know.

I would be interested as well

 

[sfx2000]

just a thought - on the iDevices (iPhone/iPad/iPod touch) - go into Settings/WiFi and on the active SSID, tap the (i) "info" and forget the network.

While still in Settings, go to General, About, Name, and change the name to something different - call it NGTest or something unique...

Do this for all iDevices on the network - giving each a unique name...

Then go back and reattach to the network, and see what happens...

The thought here is that one can have multiple devices with the same Name, but different MAC addresses, and if the DCHP daemon is binding NAME/IP rather than MACAddr/IP, then things can and will get confused...

 

[garethr]

just a thought - on the iDevices (iPhone/iPad/iPod touch) - go into Settings/WiFi and on the active SSID, tap the (i) "info" and forget the network.

While still in Settings, go to General, About, Name, and change the name to something different - call it NGTest or something unique...

Do this for all iDevices on the network - giving each a unique name...

Then go back and reattach to the network, and see what happens...

The thought here is that one can have multiple devices with the same Name, but different MAC addresses, and if the DCHP daemon is binding NAME/IP rather than MACAddr/IP, then things can and will get confused...

Nice theory although in my case all my iOS devices are already uniquely named, plus most of us with issues are only seeing this on the 5GHz band, and 2.4GHz is fine.

Should also point out that with iOS you actually don't have much control over the device's host name - it usually ends up being a 'safe' version of the friendly device name with any invalid characters/spaces removed.

 

[garethr]

Hopefully @NETGEAR Guy won't mind me posting this here, but I have a quick update on initial testing with this new beta: Good news is that for me it seems to have fixed all my iOS 5GHz connection issues. Devices connect instantly and with complete consistency to the 5Ghz band and I'm no longer seeing the multiple DHCP log entries for iOS devices either.

@Killhippie - be interested to see how you get on with testing your end as the issues you were seeing seemed to differ ever so slightly from me.

 

[sfx2000]

Nice theory although in my case all my iOS devices are already uniquely named, plus most of us with issues are only seeing this on the 5GHz band, and 2.4GHz is fine.

It was early in the morning - and I didn't have a chance to get the morning cuppa coffee

 

[Killhippie]

Hopefully @NETGEAR Guy won't mind me posting this here, but I have a quick update on initial testing with this new beta: Good news is that for me it seems to have fixed all my iOS 5GHz connection issues. Devices connect instantly and with complete consistency to the 5Ghz band and I'm no longer seeing the multiple DHCP log entries for iOS devices either.

@Killhippie - be interested to see how you get on with testing your end as the issues you were seeing seemed to differ ever so slightly from me.

Just sent Netgear Guy my info so as soon as I get it and load it I'll let you know, I hoping this is the magic bullet, even though our devices are have slightly different issues the cause I think it is similar. Especially the multiple DHCP IP issue. Also my Apple devices have as unique a name as they can have and I've never experienced this issue with Broadcom based units so maybe its a QCA/Netgear firmware issue. Fingers crossed.

 

[Killhippie]

From my testing it has solved my issues too. Mine were not quite the same as garethr but rebooting Apple devices seems fine now, also did I not see the IP fragmentation messaging on the PS4. I have reverted back to the 1.0.1.28 version as I am using some of that versions features and the reboot issue on Apple devices is not crippling for me but thanks for getting this sorted. How long before a roll out of this firmware with the improvements? Spot on for sorting this out.

 

[NETGEAR Guy]

Perfect...many thanks for taking the time to test.

Will check status with engineering and keep you posted.

 

[garethr]

@NETGEAR Guy Just wondered if you had an ETA from engineering on when we might see this iOS fix rolled into the current firmware branch? Thanks!

 

[Killhippie]

@NETGEAR Guy Just wondered if you had an ETA from engineering on when we might see this iOS fix rolled into the current firmware branch? Thanks!

There is an issue with IP Fragmentation on the PS4 using wifi on this router, I think they may be testing that as well.

 

[NETGEAR Guy]

Chasing updated FW with iOS fix and IP fragmentation issue...