[c-nsp] Latency between 3560 and ME3400

[Jeremy Parr]

I have a metro-e circuit delivered by a provider with a ME3400 as the CPE.
The provider has issued a /30, with one end being my 3560, and the other
being the ME3400. Latency is variable on the connection between the 3560 and
the ME3400, ping times can be upwards of 100ms at times, pinging from a
server on a vlan on the 3560 to the provider end of the /30. The circuit is
10mbps, but I do not know how or where it is throttled within their network.
Traffic does not need to be exceptionally high to produce this behavior,
anything over 3-4mbps will result in variably latency up to 200ms from
server to ME3400. All interfaces are negotiating at 100/full with no errors.
Flowcontrol is disabled.


Server------3560SVI---3560 Routed Port----ME3400

interface FastEthernet0/2
 description Uplink to BTC
 no switchport
 ip address 192.168.1.98 255.255.255.252
 ip access-group MetroEIn in
 load-interval 30
 no cdp enable
 spanning-tree portfast
end

interface FastEthernet0/7
 description Uplink to VMware ESXi
 switchport access vlan 1000
 switchport trunk encapsulation dot1q
 switchport mode trunk
 spanning-tree portfast
end

interface Vlan1000
 ip address xxx.xxx.15.33 255.255.255.240
end

MTR traceroute from server to outside world
 Host                                Loss%   Snt   Last   Avg  Best  Wrst
StDev
 1. xxx.xxx.15.33       0.0%   803    2.6   1.9   0.2 164.5   6.0
 2. 192.168.1.97                      0.0%   803   12.6  22.6   0.8 104.7
22.1