Wireless Access

I'm VERY new to Aruba and Wireless....so please be gentle! :)

While deploying new AP 105's at a new site, I've noticed that the Layered SNR heat map for ALL newly deployed APs show an area of reduced signal quality immediately to the top of the AP (with "top" meaning towards the top of the diagram).  I understand that the AP 105's are omnidirectional.....and also realize that many factors can impact the SNR values shown.  However - regardless of the location within the building the new APs are deployed, the "dead" space is ALWAYS showing up in the same spot for ALL APs (all are directly "above" the AP on the heat map).  

Just to play "devil's advocate", I physically spun one of the APs 180 degrees - no change.  I'm convinced that this perceived dead space is not accurate, since the building I'm working on now is rather large, and the APs are deployed in several locations (there is nothing in the building/structure that would cause reduced coverage as shown on the heat map).

Is this simply a characteristic of the Aruba OS/app?  Any ideas/suggestions would be very much appreciated!

Dan, 

Are their any other APs on this floor or the floor above or below?  The way the heatmap is rendered in the Aruba controller is based on how the "other" Aruba APs hear this AP.  If you have an AP deployed on the map that cannot hear (or hears this AP at a very low signal) that is towards the top of this map (on this floor or the floors above or below) this would explain what you are seeing.  It does not mean that the AP is not propagating in that direction well, it just means that we cannot tell whether it is or not because of the real-time information that is being provided by other APs in that direction.  This map is really accurate when you have a good density of APs (again on the same floor or above and below since this data is 3 dimensional).  Hope this helps.

Regards, 

Austin

This office in particular has several APs that have been deployed for a while....and 3 new ones that were just recently deployed.  ALL APs are showing the same basic pattern - even those that are in close proximity to other APs.

It just doesn't make sense that ALL of the "bad" spots are towards the top of the diagram regardless of the APs actual location.

Dan, 

Agree, this does look strange, have not seen this before, sorry, TAC may be able to assist if noone on the forum can.

Regards, 

Austin

How long has it been since you deployed the APs on the map?  There is a default predictive coverage that will be shown until AMP has polled the controller that terminates the AP and has gathered enough info to build a valid coverage pattern.

Oops.. sorry.  Dont know why I was thinking you were talking about Airwave.... :(

Anyway, the same is true for the controllers heat map.  You have to give it some time for the all of the information to be gathered, processed and updated before you will get valid coverage patterns.

All of the APs along the top and right side....as well as the 2 in the lower-left corner...have been there for over a year.  The 3 along the bottom center/right are only a few days old.

Is ARM enabled?  If so, do you have power-save aware scanning turned on?  If you do, ARM may be able to go off-channel and scan, so it doesn't know about the other APs.

Check the commands:

show ap monitor ap-list ap-name xxxx (where xxxx is one of the APs) - does it show the other APs around it?  Do they have good (> 30 RSSI)?

show ap arm scan-times ap-name xxxx - does Scans Attempted = Scans Rejected?

From the AP in the upper-right corner:

Some of the RSSI values are >30, some are not.  In NO instance is the scan attempts/rejects the same.

The Scans rejected should be a small percentage of the Scans attempted.  What is happening is that your APs are trying to go off channel (they do this every 10 seconds by default) to scan other channels.  They do this to try to pick the best channel.  When they attempt to go off channel, something is causing them to not be able to (that is why the Scans rejected number is so high).

In your ARM profile, do you have power save aware checked?  If so, you should turn that off (monitor closely, though, just in case you have very old clients that don't handle power save well).  That is probably what is causing the rejects.  This happens when the AP wants to scan another channel but there is at least one client in power-save mode and AP can't go scan. 

Not being able to scan also can cause performance issues due to ARM not being able to pick the best power and channel for each AP.

If I uncheck Power Save Aware Scan, could that potentially impact latency-sensitive traffic...such as VoIP, streaming video, etc.?

Not really.  If a delay sensitive app is running, the device shouldn't go into power save.  I turn power save aware scanning off everywhere I go and I have a bunch of networks running with voice, video and all kinds of sensitive data apps.