12-29-2011 08:11 AM
At the documentation I have found that the self-healing feature provides increased reliability and redundancy: the network continues to operate if an AP stops functioning or a connection fails.
I want to know how this feature works and how enable it on the controller.
Appreciate your help.
Solved! Go to Solution.
12-29-2011 09:10 AM
Are you referring to the ability for mesh to self heal around a failed link/AP or self healing of a WLAN around a failed AP and the coverage hole that may appear from that?
12-29-2011 10:21 AM
In the user guide (at least the 6.1 version I am looking at) it really only discusses "self-healing" from a Mesh perspective. That said, ARM does include self healing capabilities for the WLAN environment. While they are not explicitly called out in the user guide and their is no explicit "self healing" knob in the ARM settings, ARM will do the following by default:
- Determine amount of overlap between APs/Channels, 802.11 and non-802.11 noise in the area, and dynamically adjust channel/power accordingly, 24x7.
- If the RF environment changes due to AP failure, 802.11 or non-802.11 interference, error rates, etc, the ARM algorithm will adjust accordingly to compensate or work around the problem (like increasing power on adjacent APs to cover a hole caused by a failed AP).
To allow the above to happen, you must have ARM assignment enabled (to single band typically) and ARM scanning enabled, both of which are defaults. ARM has a bunch of other settings that determine how and when it will make a decision, best to leave these as default unless otherwise instructed by TAC or if following recommendations contained within our VRDs (for example placing ARM into an "aggressive" mode when first installing a system). One other important consideration is that the AP placement must have been designed to allow for ARM to cover a coverage hole from a failed AP. In a normal environment, APs might not be transmitting at max power because overlap was planned in beforehand and an AP would only need to boost its power if an AP fails.
This paper is a pretty good primer on how and why ARM does its thing: