Dear All, I created two vms with vpp installed and connected them with host-net. No drop was observed between the two vms with Iperf3. I then leverage tap to make the vms talk through vpp. In vm compute0: [host 123.0.1.1/24]—tap—[123.0.1.2/24, GigabitEthernet0/8/0 (up) 192.167.1.1/24 vpp] In vm compute1: [host 123.0.2.1/24]---tap---[123.0.2.2/24, GigabitEthernet0/8/0 (up): 192.167.1.2/24 vpp].
Now I can see dramatic drops after calling the cmd “iperf3 -c 123.0.2.1 -u -b 1g”: Accepted connection from 123.0.1.1, port 59798 [ 5] local 123.0.2.1 port 5201 connected to 123.0.1.1 port 57969 [ ID] Interval Transfer Bandwidth Jitter Lost/Total Datagrams [ 5] 0.00-1.01 sec 11.6 MBytes 96.5 Mbits/sec 0.088 ms 12300/13783 (89%) [ 5] 1.01-2.00 sec 16.0 MBytes 135 Mbits/sec 0.042 ms 13696/15739 (87%) [ 5] 2.00-3.00 sec 31.0 MBytes 260 Mbits/sec 0.034 ms 11529/15498 (74%) [ 5] 3.00-4.02 sec 12.9 MBytes 106 Mbits/sec 0.048 ms 13361/15009 (89%) [ 5] 4.02-5.01 sec 11.5 MBytes 97.5 Mbits/sec 0.065 ms 13925/15400 (90%) [ 5] 5.01-6.00 sec 11.2 MBytes 94.9 Mbits/sec 0.038 ms 13521/14958 (90%) [ 5] 6.00-7.00 sec 19.4 MBytes 162 Mbits/sec 0.043 ms 12867/15350 (84%) [ 5] 7.00-8.00 sec 25.7 MBytes 216 Mbits/sec 0.024 ms 12026/15312 (79%) [ 5] 8.00-9.02 sec 21.4 MBytes 177 Mbits/sec 0.035 ms 12529/15268 (82%) [ 5] 9.02-10.02 sec 17.4 MBytes 146 Mbits/sec 0.048 ms 12957/15189 (85%) [ 5] 10.02-10.04 sec 0.00 Bytes 0.00 bits/sec 0.048 ms 0/0 (-nan%) Any advice and suggestions will be greatly appreciated. Thanks, Hao
