Wireless Access

I hope someone can help.

 

Yesterday afternoon our both our guest and private networks stopped allowing 95% of internet connections.

 

We have an Aruba 620 controller with several 105 APs.  We have a guest network which has a login portal and a private network which authenticates against a RADIUS server.

 

Both of these networks remain up and can be connected to.  When using the private one, I am still able to access our servers and upload and download files to these servers without any issues.  

 

However I cannot get internet access properly using either network.  95% of web requests time out and the other 5% appear to run at modem speed.  Our wired network is still responding as normal and I have tested DNS etc.  

 

This setup has been working without issue for the past couple of years and we have made no recent changes to our network or our firewall here.

If you have not already, please open a TAC case in parallel.  There are many reasons why this would happen and we would only be guessing on this forum.

Case # 1723509

http://community.arubanetworks.com/t5/Controller-Based-WLANs/How-to-stop-unknown-unicast-from-LAN-to-be-flooded-by-the/ta-p/179934

Did this about 5 minutes ago and it hasn't made any difference.....

MickLLF,

 

Have you uploaded the logs.tar from the controller to the case?

 

 

 

I emailed them to the engineer I was dealing with earlier.  He went to take another call and advised an update of our controller.  I've since updated the controller and enabled the bcmc optimization all to no avail.  Stuck now waiting for next steps.

Please attach the logs to the case, so others will have access to it. If the logs are too big, email will block it.

Attached are the latest logs.  Thanks

 

<files removed>

Your logs contain personal information and should not be uploaded here.  They should be uploaded to the support portal at https://support.arubanetworks.com.

 

Your logs show that you rebooted the controller 7 minutes ago, so any historical issues you might have had were wiped out.  What exactly is wrong?

One of your engineers appears to have sussed it.  Still unsure why this would suddenly start causing issues when it hadn't previously however.

 

Below is the summary

 

1. We were seeing issues with clients not able to pass traffic to internet.
2. We see the issues only for wireless and not for direct wired clients on same vlan.
3. Created a new test profile and was able to pass traffic.
4. We then changed to dot1x and still able to pass traffic.
5. We then cloned to the exact production except ssid name and was able to reproduce the issue.
6. We removed the wmm config and was able to pass the traffic on test profile.
7. We did the same on production and was able to pass traffic on both ssid.
8. We believe wmm mappings are not prioritized all the way to uplink which is causing the issue