I'd be shocked if there was an "ARP bottleneck". LTE is not an ethernet-esque medium, nor is it a broadcast-capable medium (point-to-point addressing is used for every UE tunnel). The only place that ARPs enter into the equation is when the EPC "bridges" between the UE and your ethernet network. In that case, the EPC itself responds with an ARP for anyone who is asking for one from a UE with that IP address, and the ARP that the EPC sends bears the EPC's own MAC address (since UEs don't actually have MAC addresses, at least on the WAN-side).
MAC/IP pairs would need to be in your ethernet router's ARP table for any traffic destined to any UE, too, not just SNMP or ICMP traffic. -- Nathan From: [email protected] [mailto:[email protected]] On Behalf Of Jeremy Austin Sent: Monday, April 03, 2017 4:04 PM To: [email protected] Subject: Re: [Telrad] Uplink throughput again On Mon, Apr 3, 2017 at 2:29 PM, Nathan Anderson <[email protected]<mailto:[email protected]>> wrote: Have you tried shutting off only the SNMP polling for a few cycles, but leaving the ICMP polling tests running? This is why I said I've found the source, but not yet the cause. I'm going to see if I can isolate SNMP from ICMP, depending on the poller's decision tree. I'm not eliminating some kind of arp bottleneck either until I've sniffed @ the EPC. -- Jeremy Austin (907) 895-2311 office (907) 803-5422 cell [email protected]<mailto:[email protected]> Heritage NetWorks Whitestone Power & Communications Vertical Broadband, LLC Schedule a meeting: http://doodle.com/jermudgeon
_______________________________________________ Telrad mailing list [email protected] http://lists.wispa.org/mailman/listinfo/telrad
