Advanced Network Engineering Assignment
Question 1 - (Based on Netmonlab1a_16and Netmonlab1b_16exercises)
(a) (i) Explain with reference to a Wireshark packet trace, what happens when you Ping a PC on the same subnet with a packet size of 2500Bytes and what the purpose of the ping -f and ping -i switch?
(ii) The Fluke Micro-Scanner cable tester gave an out of spec result of 110m for the length of the cable between a bench-socket 28 and the hub room. If the incorrect NVP setting was 74% and the actual cable length is 96m, calculate the correct NVP value needed to calibrate the instrument.
(iii) With reference to your results from part c explain what device was most likely the performance bottleneck for Internet download speed.
(iv) Submit a screen shot of the CPU test_xl log file showing the %Processor Time and processor Queue Length counter as an Excel graph as described in lab1a Partc(ii) 7.
Explain how Performance Console was used to display a message on screen when processor activity exceeded a threshold value of 80%. Include screenshots of the message and application log for this event.
(b) (i) Submit the completed Table 1 from part b 4 and comment the effect of TCP receive window size (RWIN) on throughput performance for each PHY. Estimate the min RWIN size for Gigabit Ethernet to get maximum throughput for RTT = 0.1ms.
(ii) Report on the streaming video performance investigation over the 10BaseT Ethernet network you undertook in Part 2 (b) of lab1b. Include any relevant results to support your conclusions.
(iii) Include a Wireshark decode to confirm that TCP checksum offloading was being done in part c (i) 6 and explain why it increased throughput performance in the benchmark tests.
(iv) Discuss the theoretical reasons for employing Jumbo frames in the big file download scenarios and account for any unexpected results you obtained in practice.
Submit Wireshark results to show that Jumbo frames were actually being used and explain why ping -l 2500 -f doesn't cause frame fragmentation when Jumbo frames are transmitted.
Question 2 - (Based on netmonlab2_16exercise)
(a) (i) State the information you obtained and OID for the Agent PC when you issued the Get command from the MIB browser for the SysDescr object.
(ii) Submit your Wireshark SNMP packet decodes of the following:
- Response packet carrying the SysDescr information highlight community name, SNMP version and UDP port number.
- First trap generated when the SNMP agent PC service was restarted, highlighting the trap type and UDP port number.
(b) (i) State the CLI commands used to configure the 2960 switch for operation in the public community with read only access and trap destination 192.168.1.101.
(ii) List the other management options for configuring SNMP on the Cisco 2960 switch.
(ii) Explain how you would configure the Cisco2960 switch so that activity on port 1 can be monitored by a monitor PC connected to port 8.
(c) (i) Submit the NetworkView map you obtained in Part4a showing the devices discovered on you subnet.
(ii) Show the SNMP info discovered by NetworkView about the Cisco router and list the non SNMP discovery methods it uses to obtain information about network devices.
(iii) State the purpose of the Cisco CDP and Microsoft LLTDP protocols. Submit Wireshark decodes showing the content of a CDP and LLTDP frame and a screenshot of the network maps.
(iv) Write a script which can execute as a batch file on a Windows computer to discover the IP address, MAC address and NetBIOS names of all the devices on a subnet and save the results to a .txt file located on the desktop.
Assume that you know the subnet number; that all the devices have correct addresses/subnet masks and respond to ICMP requests.
Submit test results showing at least two discovered devices.
Question 3 - (Based on NetmonLab3_16 exercise)
(a) (i) Submit the table of results from part A showing the wireless networks detected with the InSSider site survey tools. Define the RSSI and S/N metrics used in the table.
(ii) Calculate the power level in nW received by your network adapter and the noise level in dBm for a RSSI of -60dBm and S/N of 30dB.
(iii) Review the WiSPY spectrum analyser recording made during the lab session (see BB Support Resources Folder) and then identify the first time the video sender was used to jam the nwlan-5 access point. Submit a screen shot showing this period.
Suggest how video sender interference could be minimised for 802.11b/g mode and why 802.11a or 802.11ac mode might be a better choice in this case.
(b) (i) Submit your TCP throughput results from section D Table 2 along with an Excel histogram chart to help you compare the performance of the different scenarios.
Confirm or otherwise the expectations that IEEE 802.11n provides 6x the throughput and performs better than legacy IEEE 802.11a/b/g networks in mixed mode operation.
(ii) Explain why your data throughput results are much less than the raw transmission speeds for each wireless mode and account for the higher throughput between a wireless station and wired server compared to wireless to wireless station.
(iii) Describe how multiple spatial streams, 40GHz channels and short guard interval used by IEEE 802.11n equipment increases wireless transmission speed.
Locate the MCS index for the 3x3 to 2x3 MIMO system used in the lab and state the modulation technique and best link speed(data rate) it can employ. Use the values in the table to show how the link speed is determined for this particular index.
(c) (i) With reference to the Wireshark VoIP call Statistics flowgraphderived from the packet capture, explain the 3 phases of a SIP phone call.
(iv) Submit one of the decoded RTP packets from your VoIP tests highlighting the ITU G.711 codec payload and RTP, UDP and IP headers. Determine the voice payload size and show that it corresponds to the default G.711 20ms sample size in bytes. Calculate the bandwidth requirement for a G.711VoIP call over a LAN for a frame overhead of 18 bytes.
(v) Present the graphical results of the VoIP QoS tests made with the Expert Analysisfeature of Observer and indicate the R-Factor and MOS score obtained. Confirm or otherwise the expectation that IEEE802.11 supports reliable VoIP even when sharing the network with heavy data loads.
Question 4 - (Based on netmonlab4_16 exercise)
(a) (i) Calculate the theoretical minimum time to transfer a 100MB file over a TI WAN link using FTP and explain why it may take much longer in practice and why upgrading to a T3 service may not improve throughput very much for a long distance link.
(ii) Submit a comparison graph showing the results you obtained for the RiverbedIT GURU WAN TCP window size optimisation simulation lab for the file transfer over a T1and T3 link for default 8KB window size and when the window size is increased to 65KB.
(iii) Compare the simulation and emulation methods for testing this scenario and list the pros and cons of each.
(b) The SG TCP Optimizer utility below allowed you to view and tune TCP settings for your client PC.
(i) Distinguish between the normal and disabled TCP auto tuning levels and explain why for MSS=1460 Bytes the RWIN value of 64240Bytesmay give better throughput performance than RWIN=65535 Bytes.
(ii) Show screenshots of the SG TCP/IP Analyzer results for normal and disabled TCP auto tuning levels for the Downlink Global Speed test and submit Wireshark decoded TCP SYN packets for each case with the following details highlighted:
- TCP window size
- MSS
- Scale factor
- Sack status
(iii) State the max RWIN value achieved for the normal level download speed test and a screenshot of a TCP ACK packet advertising this value.
(c) (i) Submit in one table the results for the real and emulated BT broadband speed tests.
(ii) Discuss your findings for the broadband emulation and account for any major differences between the real and emulated results.