Wireless Access

I have about 75 deployed RAP-2 and a 3 RAP-5s deployed in my environment. The RAP-2 are in users homes and the RAP-5 are in remote offices. Been receiving complaints that the RAP-5 were disconnecting users for short periods of time. These three RAP-5s all use Comcast as the ISP. I'm also using Airwave. I have the SNMP poll set to 5 min. Sometimes I don't get notifications that the RAP 5 went down/back up. I think this is because the RAP will complete the boot before the 5 minute interval.

I enabled an SNMP trigger to email me when the message "wlsxAccessPointIsDown" or "wlsxAccessPointIsUp" is received. Last night I noticed about 21 RAP2s went down. 17 of which are on Comcast. I am debating increasing the Bootstrap threshold from the default of 8 to 16. 

Has anyone seen an issue like this and is increasing the boostrap threshold the right thing to do?

Thanks

First things first:  What version of ArubaOS?

That would help..sorry for not including it. On version 6.1.2.3. 

The bootstrap threshold for RAPs is 30 by default, even though the GUI shows 8 it is 30 by default for RAPs, the 8 on the GUI applies to campus APs, for RAPs only a value higher the 30 will trigger a change of the defualt value of 30 to a higher value. If is highly recomended that you check the BW and latency on these links. Also make sure that your WAN on the controller side is not suffering from any BW and lantency issues.

Regards,

Sathya

I'm not sure if this is true. I have it set to 8 in the gui, but a "show ap ap-config ap-name" also shows a value of 8 and not 30. I've attached a screenshot.

Also want to add that this is the output from a "show ap debug counters ap-name". The link is a 12 Meg Comcast Business connection. The controller has a 100meg ethernet connection from Level3. I truely think it's an issue with Comcast because I'm seeing other (RAP-2 that users have at home using comcast) will disconnect as well. Not all at the same time but it's not uncommon to see 2 or three of them go down at the same time. 

Rebootstrap Information
-----------------------
Date       Time     Reason (Latest 10)
--------------------------------------
2012-02-02 11:17:40 Missed heartbeats
2012-02-02 11:29:03 Changing to LMS #0 (64.30.17.70)
2012-02-02 17:38:09 Missed heartbeats
2012-02-02 17:38:19 Changing to LMS #0 (64.30.17.70)
2012-02-02 19:28:31 Missed heartbeats
2012-02-02 19:30:05 Changing to LMS #0 (64.30.17.70)
2012-02-03 05:46:07 Missed heartbeats
2012-02-03 05:47:19 Changing to LMS #0 (64.30.17.70)
2012-02-03 06:48:21 Missed heartbeats
2012-02-03 06:49:34 Changing to LMS #0 (64.30.17.70)

---

@tbryant wrote:

That would help..sorry for not including it. On version 6.1.2.3. 

---

The best thing you can do is to setup a syslog server so that you can see what it happening at the time you have an incident.  You may want to determine if certain APs disconnect (individual issue) or if all APs disconnect or have issues (controller or headend issue).  

As of ArubaOS 6.x, ipsec retries (by default 360) are used to determine how many times an AP will try to connect to a controller before rebooting, if it loses connectivity.  You will see those messages in the system log when the APs attempt to reconnect.