Wired Intelligent Edge

Hello all,

I've posted to this board recently with some logs from our controller and they weren't very helpful.  Let me explain my scenario here.

I had a network consisting of the following.

1- 3600 -master

1 7210 - local

3 - 3400 locals.

The 7210 was added as a local and then I began to terminate all APs to it for performance.  It got to the point where I could no longer upgrade my IOS because the 3600 would not go past 6.4.  I recently removed the 3600 master and made the 7210 a master.  I changed the DNS record for aruba-master and all is working great.  All the AP's were already tunneled through the 7210.  So traffic was not interupted very much.  All 292 AP's came up and traffic was flowing like it was before.

The reason I wanted to do this upgrade was because we had some latency issues on the network(thinking we just needed to upgrade to address these issues.)  Dropped packets randomly on the wireless.  When coverage was not an issue.  It didn't matter where we were on the network.  Traffic was slower than it should have been.  We just put in a new Aruba HPE 5400 chassis and uplinked two 10 GB SFPs to that core.  So bandwidth didn't seem to be the issue.  We upgraded to the latest OS, hoping that all would be good.

Now I've still got the performance issues, but also certain Macbook Airs(early 2013) are randomly dropping network access(Still connected to wireless and associated, but dropping packets completely)

I've opened up a TAC case and went to Level 2.  Everytime I tell them my issues, they begin to evaluate the wireless for radio interference, etc.  The one tech asked if there was a cell phone tower nearby.  We did client debugs and the clients did not drop association with the AP, nor did they roam. However, they still dropped connection.  We have almost 3000 devices on the network and a handful of Macbook Airs now losing this connection and others were still slow and dropping packets.

We then decided to change the APs and put them in bridging mode on the VAPs.  This gave us a good boost in performance, network wise.  Page loads much faster. speed tests are more what they should be.  Those few Macbook Airs are still dropping connection randomly but keeping association.

Along with this issue, after switching to bridging mode and tagging the VLANS on the AP ports on our Aruba/HP network.  We are seeing our HP printers losing connection on the wired network, Until we change them to a new VLAN and then they keep their connection and we can continue to ping them.  The printers are on the save VLAN as the wireless clients.

Needless to say, we are chasing one little issue after another. One thing that has improved is our connection speed for those clients that are working since we moved to Bridged Mode.  

I know this is a lot but TAC, Tier 1 or 2, doesn't seem to understand what types of issues we are facing and timliness and always having to create logs is cumbersome when we need efficiency and these issues fixed NOW. 

Any thoughts on these issues?  I'm a novice when it comes to most of this.

What is the TAC case number?

I will say that it is very difficult to troubleshoot issues remotely.  TAC is just asking questions, guessing what the problem(s) could be.  Unfortunately, there is almost no limit to the number of things that can cause casual disconnects, so many questions have to be asked to gain an understanding of the state of your network.  It is also very difficult, if not impossible to understand the state of a client just from controller logs.

Long story short, you should enable user debugging for those clients that have had historical issues.  When the problem starts, you can obtain the logs.tar with tech support from the GUI of the controller so that the TAC engineer can observe the state of the controller, as well as any of the clues that the client may give from the logs.  That will allow the TAC engineer to narrow down the unending list of things that could be causing your issue.