Roaming for wireless clients is very important. Some smaller environments pay little attention to roaming; while some environments require near perfect roaming and can’t afford for wireless clients to drop a connection and then reconnect. Wireless engineers spend lots of time figuring out how to deploy a wireless network that will offer seamless roaming. And sometimes we have to assess an existing network where roaming is an issue. Roaming troubleshooting is an important troubleshooting skill; it is also very difficult because you need to capture on multiple channels and try to replicate the problem and hope that you can capture the right data.
Let’s say you could do all that and did not see any issues with the roaming, or maybe the intermittent issue resolved on its own, or you don’t even know what exactly happened when user(s) complained and you need to first define and identify the problem?
As a network and wireless engineer, we are all used to getting escalated tickets of roaming issues and disconnects, but we have always had to struggle with the identifying and defining the issues. Going back in time and looking at the roaming experience of a device a day or two ago has always been a challenge.
Mist – Insights
Mist Insights and Client Insights gives ability to do just that. I can go back in time and look at the Client Insights, understand their roaming behavior and co-relate the data to Identify the issue.
I will use this one client that seems to be having issues with connectivity and roaming. First under Client Insights, I changed the range to “Yesterday”. This will enable me to look at the data from yesterday because I need to look at the issues in the past, specifically the roaming issues.
Scrolling down, towards the bottom of the page is where I want to be
- Post Connection: Shows RSSI values during different times
- TX/RX Bytes: Shows Tx/Rx Bytes during different times
- Roaming: Shows the roaming behavior of a client during different times.
- All these three things can be co-related. Look at the brief clip.
As I move my cursor through one section, I can see that it is also moving in the other two sections. I don’t have to figure out the time frame for each data point. Mist AI is doing that for me. For example, looking at the screenshot below and all the red arrows. I can see what the min/max RSSI values were at a particular time, what the Tx/Rx Byte values were and what was going on with roaming (in this case, my client is not having a good day).
I can dig deeper into the roaming by clicking on the Roaming window. This view can provide me additional details:
- Starting AP
- Roamed AP
- BSSID
- Protocol
- Channel
- Frequency Band – Is the roam between 5 GHz to 5 GHz, 2.4 GHz to 2.4 GHz, 5 GHz to 2.4 GHz or 2.4 GHz to 5 GHz
- RSSI Values – What was the RSSI before roaming and what is the value after roaming
There are lots of data points here and more issues and roaming events will add more data in here. I can easily zoom in and view all the data points clearly.
Final Thoughts
Ability to go back in time and look at unique data points/co-relate the information is extremely helpful when trying to identify the problem. I can see where is the client having issues, what are the RSSI values as he roams between APs, is there are pattern or just onetime thing, certain events will even have frame captures attached which can further help with the analysis of the issue. All this, from my desk.