@msabin wrote:
Active survey. Which explains the occasionally "missing" AP.
You might want to share your thoughts on the load-balancing features with the Engineering Team, whenever I call in for support that's the first thing they want turned on. I have a very sparse environment, 90,000sqft - 25 AP92/93 - 30 clients.
msabin,
What I mentioned is an approach based on limited information, not hard and fast rules. I do not know the details of your deployment and it is difficult for anybody who is not onsite to also know that. As a result, I cannot say what engineering say is bad, I can only offer a general approach based on what I see here. If you call in to engineering and you say you have a roaming issue, they might tell you to turn on load balancing or clientmatch to solve your specific issue, and it might work.
If you want to survey to diagnose a problem, I would first do a passive survey without those features turned on to see the true power levels a client is getting in all areas to see if it would better help to supplement coverage, rather than turn those features on. Load balancing features can only do so much to make up for poor coverage and a poor client roaming algorithm. Baseline your coverage first, to get a good general impression of how it looks. Make sure you do a side-by-side with the client in question to see the difference between the laptop adapter you are measuring and how the client reports. Contact the manufacturer to see if they have any advice on the roaming threshold and see if you are getting that in the passive survey.
Everything that I am mentioning might seem overly basic, but that is what is necessary before any load balancing mechanisms are turned on. Again, RF features like Client match work the best when basic coverage is met, and can only do so much...
In the end, if you intend to do an active survey, you have to configure your survey laptop's adapter to match as many characteristics of the client so you can be ensured that is the way the client will truly operate in that environment.