We recently saw something similar to this.  Our Cisco WLC showed a message like this:   Thu Sep 24 15:16:54 2015 RADIUS server 1.1.1.1:1812 failed to respond to request (ID 217) for client f8:95:c7:a6:34:7c / user 'f8-95-c7-a6-34-7c'   We found the LG Phones were sending an illegal EAP code of 53 which was being silently ignored by our authentication server.  According to RFC3748 the WLC should also drop this and not pass it to the authenticator in the first place. Because the Auth server was silently ignoring the request, the WLC thought the auth server was dead and tried to switch to another one.  The the WLC keeps flipping between auth servers, which also cuts off other client authentications that are in progress.  This actually causes queuing issues on the auth server as well because it is waiting on the requests that got cut off by the WLC switching to another auth server. Right now we are Blacklisting the MAC's of these phones as a workaround.  We need a more permanent solution of course.  It may be a while before the WLC follows the RFC and discards EAP frames with illegal codes. Any suggestions?   ... View more

But since there is only one match statement in this case, wouldn't match-any be equivelent to match-all? on the other had, if the class map looked like this I understand how match-any would be differant than match-all class-map type qos match-all VIDEO      match dscp af41      match dscp cs4 i.e. In this case I would expect no packets to ever match since a packet could never have dscp af41 AND dscp cs4 But with the original class-map class-map type qos match-all VIDEO      match dscp af41 cs4 Can I assume an OR is applied between "af41 cs4" since it is a single match statement? Brian Holmes Verizon Enterprise for Chrysler ... View more

The 5K appears to set the bandwidth to 0 if I do not set the bandwidth: N5KZTL01# show queuing interface e1/1 Ethernet1/1 queuing information:   TX Queuing     qos-group  sched-type  oper-bandwidth         0       WRR             55         2       WRR              5         3       WRR             40         4       priority         0 Should I be concerned? Brian Holmes Verizon Enterprise for Chrysler ... View more

One other thing I had to send was TACACSPLUS-Priv-Level = ROOT which by the way was not in any manual.   ... View more

I notice the one vendor has: random-detect exponential-weighting-constant 9 while the other has random-detect exponential-weighting-constant 0 How might this relate to the latency increase during saturation? ... View more

Thanks all for the feedback. We have the carrier engaged, but they are giving the standard "it is working per design" answer.  From reading a QOS book from Cisco Press, they say that WRED should be used in the Core of the network and not the edge.  Would a CE router be considered Core or Edge?  It think the E stands for Edge but maybe they mean the user edge. ... View more

This is a 6Mb Multilink interface and the reserved bandwidth for AF31 is no where near exhausted even during circuit saturation. The carrier is stating that WRED is being performed before the packets hit the custom queue which seems like the reverse of how things should be handled. Sent from Cisco Technical Support iPhone App ... View more