Reached Buffer limit

[farzan_qureshi_]

Hi there,

We are using H510s at a customer site. I see following in the AP syslog:

Dec 13 05:10:12 kern.warn kernel: [84170.752918] Reached buffer limit
Dec 13 05:10:12  kern.warn kernel: [84170.755246] Reached buffer limit
Dec 13 05:10:12  kern.warn kernel: [84170.757607] Reached buffer limit
Dec 13 05:10:12  kern.warn kernel: [84170.763184] Reached buffer limit
Dec 13 05:10:12  kern.warn kernel: [84170.765507] Reached buffer limit
Dec 13 05:10:12  kern.warn kernel: [84170.767782] Reached buffer limit

Following is the version command output:


Ruckus H510 Multimedia Hotzone Wireless AP
Version: 5.1.1.0.624

The APs are managed by Vsz 5.1.1.0.5985.1.0

Any ideas what may be the reason? We are getting complaints from the users that they are not able to associate to AP.

[farzan_qureshi_]

So for us it got resolved when we moved from version 5.1 to 5.2. We don't see any buffer overflow errors. I am not sure about Unleashed version. It they are already on latest version then nothing that you can do other than log it with Ruckus Support.

[paul_johnson]

This seems to be the pattern of logs;

kernel: ol_rc_update_handler: WARNING: handler called without transmit attempts

wifi0: 63661671 : stuck beacon (missed 3) scanning 0, syncbeacon 0, sc_beacons 1,sc_isCaldone: 1

12:18:11 ZD-APMgr: IPC_thread rcv ping from TACMON 
12:17:56 kernel: Reached buffer limit
12:17:56 syslog: bba: STAs from 2.4g to 5g.
12:17:56 syslog: bba: sta-2g=6 sta-5g=1 perc=50 sta-2g-expect=3

11:17:56 kernel: Reached buffer limit
11:17:56 syslog: bba: STAs from 2.4g to 5g.
11:17:56 syslog: bba: sta-2g=6 sta-5g=3 perc=50 sta-2g-expect=4

[syamantakomer]

Hi Paul,

Is there any service impact of this?

[paul_johnson]

There are service impacts at the site and we are trying to narrow down the cause, we are not certain this has anything to do with it but buffer limit reached and stuck beacons stood out in the logs.

Clients complain of service degradation, two VPN users complain throughout the day performance gets worse and very slow speed (daily traffic about 400-700mb each of these two clients) . Also intermittent iPhone "unable to connect to network" with correct credentials and good RSSI.

5-15 clients max, load and band balancing is off

[farzan_qureshi_]

With iPhones, they have introduced random mac address feature. Try to disable it and then see how it goes.

Also, disable RSSI threshold if you have enabled on the APs where a device should be able to communicate with RSSI of -67 or something stronger to get a response from the AP.

Also, if you have 80 Mhz channels enabled, please decrease the bw to 40 Mhz channel or even 20 Mhz if the data requirements are not that high. Honestly, 20 Mhz will serve the purpose well and client devices will have stable connection health. However, placement of APs also matter.

I am sure you have already checked for the channel overlap in the premises. 

Also, if channel fly is enabled, please move to background scanning and increase the scan interval to 3600sec.

Let us know how it goes.