This assignment was written by Parsa Tajik in December of 2022.
While in the assignment's directory, enter the following commands in your terminal to clear mininet and start the testing process:
sudo mn -c
sudo ./run.sh
1. Why do you see a difference in webpage fetch times with small and large router buffers?
Q20
Latency Average: 0.589277777778
Latency Standard Deviation: 0.214816279529
Q100
Latency Average: 1.33963333333
Latency Standard Deviation: 0.695737904834
Our varied fetch times are a result of different buffer sizes and TCP's packet loss detection algorithm. When travelling from h1 to h2, the packets are bottlenecked by the slower link and forced to wait in the buffer. TCP continues to increase CWND and send more packets until a loss is detected and window size is modified. However, this only occurs when the buffer is full. Therefore, when the buffer size is larger, more packets are going to be queued, resulting in longer fetch times. Ultimately, congestion is better managed by utilizing smaller buffers.
2. Bufferbloat can occur in other places such as your network interface card (NIC). Check the output of ifconfig eth0 on your VirtualBox VM. What is the (maximum) transmit queue length on the network interface reported by ifconfig? For this queue size and a draining rate of 100 Mbps, what is the maximum time a packet might wait in the queue before it leaves the NIC?
eth0's ifconfig output:
eth0 Link encap:Ethernet HWaddr 08:00:27:ea:8e:68
inet addr:10.0.2.15 Bcast:10.0.2.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:47974 errors:0 dropped:0 overruns:0 frame:0
TX packets:44169 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:3936328 (3.9 MB) TX bytes:74442569 (74.4 MB)
Hence, the maximum transmit queue length (txqueuelen) is 1000 and the max transit unit (MTU) is 1500. Ultimately, the maximum packet queue wait time is:
Max Queue Wait Time
= (1000 packets * 1500 bytes * 8 bits/byte) / 100Mbps
= (1000 * 1500 * 8) / (10^8) s
= 0.12 s
3.How does the RTT reported by ping vary with the queue size? Write a symbolic equation to describe the relation between the two (ignore computation overheads in ping that might affect the final result).
Average RTT per Queue Size:
20 -> 36.304s
100 -> 213.204s
K = (213.204-36.304) / (100-20) = ~2.21 (constant/slope based on our graphs)
RTT = Queue Size * K
RTT is directly correlated with Queue Size
4.Identify and describe two ways to mitigate the bufferbloat problem.
-
Utilizing queue management algorithms for fixing the drop tail problem in case of bursty flows by dropping packets before the queue gets full.
-
According to our experiment, we can also reduce the maximum buffer size to decrease the overall latency. It seems as though having multiple smaller buffers is a great way to mitigate the bufferbloat problem.