I suppose that you have a MEC between you distribution layer and your VSS, right? Anyway, if from what I am understanding your SPAN destination is a link of chassis A of your VSS there is no need to add anything else on chassis B. Traffic in vlan 10 will traverse the VSL link from chassi B to reach your destination port on chassis A. This is something you should be aware of anyway... you might congest the VSL link in this case (the locality rule prevent this from happening on normal circumstances; traffic leaving a VSS MEC will always use the local leg(s) of the MEC unless it is down). Make sure you have enough bandwidth on your VSL link. Riccardo ... View more

Hi Roman, it is explained on the same page your link refers to: By default, the 10/100 ports and the dual-purpose ports on Cisco ME 3400-12CS and ME 3400-2CS switches are configured as UNIs, and the SFP-only module uplink ports are configured as NNIs. No ports are ENIs by default. If the switch is running the metro base or metro access image, only four ports on the switch can be configured as NNIs at one time. If the switch is running the metro IP access image, there is no limit to the number of NNIs that can be configured on the switch. All ports on the switch can be configured as UNIs or ENIs. The default state for a UNI or ENI is administratively down to prevent unauthorized users from gaining access to other ports as you configure the switch. Traffic is not switched between these ports, and all arriving traffic at UNIs or ENIs must leave on NNIs to prevent a user from gaining access to another user's private network. If it is appropriate for two or more UNIs or ENIs to exchange traffic within the switch, the UNIs and ENIs can be assigned to a community VLAN. SeeChapter 11 "Configuring VLANs," for instructions on how to configure community VLANs. Note Even though the default state for a UNI or ENI is shutdown, entering the default interface interface-id command changes the port to the enabled state. The default status for an NNI is administratively up to allow a service provider remote access to the switch during initial configuration. A port can be reconfigured from UNI to NNI or ENI and the reverse. When a port is reconfigured as another interface type, it inherits all the characteristics of that interface type. When you reconfigure a UNI or ENI to be an NNI, you must enable the port before it becomes active. Changing the port type from UNI to ENI does not affect the administrative state of the port. If the UNI status is shut down, it remains shut down when reconfigured as an ENI; if the port is in a no shutdown state, it remains in the no shutdown state. At any time, all ports on the Cisco ME switch are either UNI, NNI, or ENI. I think what you are looking for is there. Riccardo ... View more

you are welcome. please rate and close the question if useful Riccardo ... View more

can you span fa0/4 and fa0/6 on the cat2960 (making sure you capture vlan tags) while you ping end to end from your laptops?  so we can sure that the double tag imposed by huaweii switch is compliant with specs. Riccardo ... View more

replied in mpls section ... View more

Hi there, no the SIP400 cannot be configured as L2 switchport (it does not make any difference which SPA you actually use). However you can have multipoint Layer 2 bridging service by using bridge-domain under service instance configuration (EVC) something like: interface GigabitEthernet5/1/1 service instance 1 ethernet   encapsulation dot1q 20   bridge-domain 20 Internally one bridge-domain is mapped to one of the 4K global VLAN resource; meaning that you still need a SVI for the L3 termination. i.e: Interfaca vlan 20 ip address 10.0.0.0 255.255.255.0 regards, Riccardo ... View more

routers don't care of L2 control plane PDU's (among which you have BPDU's) as they don't process them (nor route them). So no particular mechanism to make sure they are not dropped. On the other hand swithes care of them and they make sure they are correctly sent to the CPU to be processed. The way they do is that as soon as the hardware plane see frames destined to those well known MAC addresses it 'punts' them to the CPU right away bypassying any other forwarding or traffic handling logic. Riccardo ... View more

Hi Richard, SPD mechanism is there to prevent that also routing protocol packets (or control plane ones in general) are dropped if the interface is congested. For details please see: http://www.cisco.com/en/US/products/hw/routers/ps167/products_tech_note09186a008012fb87.shtml regards, Riccardo ... View more

Hi, for such short messages don't attach outputs or few people will take the time to download and unzip the file. Error %SNMP-3-INPUT_QFULL_ERR: Packet dropped due to input queue full this message simply says that a given snmp packet was not sent to the CPU, which is supposed to receive and process such packets, because the CPU queue dedicated for the purpose is currently full and the packet is consequently dropped. Usually the message can be simply ignored as SNMP is treated with low priority and no serious issue occur if snmp packets are dropped. also the design of this device for which snmp packets end up in a dedicated queue prevents the cpu to be overwhelmed by the flooding of such packets. however if you often see those messages maybe it is time to investigate a little. reasons for the message may be (not all possible reasons are listed): - the nms console(s) are requesting too much information from this switch which cannot cope with the rate > verify if the request rate is the expected and in case slow it down. - some kind of network issue (l2 loop mainly) is causing the augmented snmp rate. verify if something in that sense is happening and rectify. - the cpu utilization is high on its own and the device is starting to drop packets sent to it starting from the ones with lower priority. > verify if this is case case and understand why CPU is high. Riccardo ... View more