Wireless Access

Just putting the feelers out there to see if anyone else is having issues with this release?

 

We have 3 main issues.

Captive portal redirects are randomly failing for IOS users, they can not get to the CP landing page, even if I enter the URL directly. Other users attached to the same Ap have no issues. Give it a little while and then they are fine, and someone else has the issue who was previously fine.

 

IOS/Apple users just suddenly appear to lose connectivity. Actually they are associated, but their speeds are so slow, it is impossible to load any web pages. This behaviour will carry on accross AP's when they roam, and then just as suddenly, normal service is resumed and they are fine again. Other users in the ame are unnaffected. 

 

We are getting lots of complaints from Apple users over general wireless performance, as ever with no detail, however Im finding if I disable band steering for the users, their performance improves dramatically.

 

The only thing that is common is that for IOS users who complain, they are all on 12.1.4.

 

I do have a TAC case open, who have been excellent, but due to the highly random nature of who is affected, it is proving a nightmare to troubleshoot.

 

As a fix we are going to move to 8.4.0 to see if that is any better (40 pages of bug fixes in the release notes!) and I have a couple of questions?

Is 8.4.0 the very latest release?

I remember there was a build of 8.2 that was considered very stable - can anyone advise on which version (that is my plan B)

The latest conservative release is 8.2.2.4 (not 8.2.4, that does not exist).

 

I would recommend 8.3.0.6 before going to 8.4.0.0 at this point, unless you need functionality specificly in the 8.4 release. I'm not aware of iOS specific issues with 8.3.0.4, so would recommend working with TAC before upgrading firmware to avoid adding new variables that could further delay getting to the root cause of your issue(s).

Do you have "drop broadcasts and unknown multicast" enabled on all of your virtual APS?

Are you running 20/40 or 80 mhz wide channels?

Sorry Charlie, I have corrected my post. I have a phonecall with TAC tomorrow to discuss further, but as this is causing issues for so many different users at random points of time, we have to be seen to do something.

I have an test IPad that I have been walkign around with, 90% of the time it is fine, 10% of the time unusable with no pattern for time or area where issues occur. That is how random this is.

 

 

cjoseph

 

I can see a Drop Broadcast and Multicast: option which is unticked. I cant see any other option?

We allow 20/40 or 80 mhz wide channels, I was advised to just let the Aruba's do their thing when we upgraded to AOS8, we run airmatch and we don't configure it in any way. To be fair, the load balancing of clients accross AP's has been superb in AOS8, which is partly why we have left things alone.

Random broadcast traffic can be a source of traffic degradation. You should try to check the Drop Broadcasts checkbox on all your virtual APs to see if it fixes your issue.

Letting airmatch determine the channel width can be good if you have no performance issues. If you have performance issues, it is best to consider restricting it to 40 mhz or below.

With that being said you should only change one thing at a time.

mmacleod did you end up resolving the issue and if so how?

We are experiencing similar Apple disconnects with 8.3.0.3.