π Lecture 85 β Next 5 Common Network Problems
π Overview: This lecture explores five additional common network problems beyond the initial four. It covers route redirections, along with packet loss, packet corruption, retransmission issues, and network congestion, explaining their causes, symptoms, and impact on network performance.
ποΈ Topics Covered
This lecture begins with route redirections, explaining how ICMP redirects help hosts find better paths. It then covers packet loss and its causes, packet corruption and detection methods, retransmission issues and the TCP retransmission timer, and finally network congestion including TCP's congestion control mechanisms.
π Lecture Summary
Route Redirections
When a host sends packets to one local router but another local router has a preferred path, a route redirection may occur. The receiving router responds with an ICMP packet containing the IP address of the recommended router. This is rare because most networks have only one router. π‘ Why this matters: Route redirections help optimize network paths dynamically but can indicate misconfigurations.
π Definition β Route Redirection: A network mechanism where a router informs a host that a better path exists through a different router. π Example: Host A sends packets to Router X, but Router Y has a better path to the destination. Router X sends an ICMP Type 5 packet with Code 0 (Redirect for Host) or Code 1 (Redirect for Network) to redirect traffic.
Packet Loss
Packet loss occurs when data packets fail to reach their destination. Common causes include network congestion, faulty hardware (e.g., bad cables, failing routers), wireless interference, and buffer overflows. TCP detects packet loss through timeouts or duplicate ACKs and responds by retransmitting lost packets and reducing its transmission rate.
π Definition β Packet Loss: The failure of one or more transmitted packets to arrive at their destination. π Example: During heavy network congestion, a router's buffer may overflow, causing it to drop incoming packets. The sending host detects this through the absence of expected acknowledgments.
Packet Corruption
Packet corruption happens when data bits are altered during transmission, usually due to electrical interference, faulty cables, or signal degradation. TCP uses checksums to detect corruption. When a corrupted packet is received, the receiver discards it and does not send an ACK, causing the sender to retransmit.
π Definition β Packet Corruption: An error where bits in a packet are changed during transmission, making the data invalid. π Formula: CRC or checksum calculation β verifies packet integrity. π Example: A noisy electrical cable causes bit flips in a packet. The receiver computes the checksum, finds it doesn't match the transmitted value, and discards the packet. No ACK is sent, so the sender retransmits.
Retransmission
Retransmission is TCP's mechanism for recovering lost or corrupted data. When a sender doesn't receive an ACK within a certain time (the Retransmission Timeout or RTO), it resends the packet. TCP also uses fast retransmit when it receives three duplicate ACKs for the same sequence number.
π Definition β Retransmission Timeout (RTO): The time a sender waits for an ACK before assuming a packet is lost and retransmitting it. π Example: Host A sends packet 5. It sets RTO to 200ms. If no ACK for packet 5 arrives within 200ms, Host A retransmits packet 5. If three duplicate ACKs for packet 4 arrive (indicating packet 5 was lost), Host A immediately retransmits packet 5 without waiting for RTO (fast retransmit).
Network Congestion
Network congestion occurs when demand exceeds capacity on network links or routers. TCP uses congestion control algorithms like TCP Reno and TCP Cubic to manage this. When congestion is detected (via packet loss), TCP reduces its congestion window (cwnd) to decrease the sending rate. TCP's congestion control includes slow start, congestion avoidance, fast retransmit, and fast recovery.
π Definition β Congestion Window (cwnd): A TCP sender-side limit on the amount of data that can be sent before receiving an ACK, adjusted dynamically based on network conditions. π Formula: cwnd determined by slow start (exponential growth) and congestion avoidance (additive increase, multiplicative decrease). π Example: During slow start, cwnd starts at 1 MSS and doubles every RTT (1, 2, 4, 8...) until a packet loss is detected. Upon loss detection, TCP Reno sets cwnd to half its current value (multiplicative decrease) and enters congestion avoidance (additive increase by 1 MSS per RTT).
β Key Takeaways
Route redirections involve ICMP Type 5 packets directing hosts to better paths, though they are rare in simple networks. Packet loss and corruption are detected through missing ACKs and checksum failures respectively, triggering retransmissions. Retransmission timing is governed by the RTO, which adapts based on measured RTT. Network congestion is managed by TCP's congestion window through algorithms like slow start and congestion avoidance. Understanding these problems is critical for diagnosing network performance issues and optimizing TCP behavior.
π§ Quick Revision Questions
- What ICMP packet type indicates a route redirection is occurring, and what codes are used?
- How does TCP detect packet loss, and what two mechanisms does it use for retransmission?
- What is the purpose of checksums in packet transmission, and what happens when corruption is detected?
- Explain the difference between TCP's slow start and congestion avoidance phases in terms of window growth.
- What is the Retransmission Timeout (RTO), and how does it affect data transmission?
π Lecture 86 β Next 8 Common Network Problems
π Overview: This lecture examines eight additional common network problems that network analysts must diagnose. It covers storms, loops, malware, name resolution failures, addressing errors, TCP inefficiencies, client misconfigurations, and application performance issues, providing practical symptoms for each.
ποΈ Topics Covered
The lecture covers broadcast/multicast storms, switch loops, virus/malware on hosts, network name resolution issues, network address resolution problems, hardware address resolution failures, lack of Selective ACK support, absence of Window Scaling, client misconfiguration, low MTU size, TCP port number reuse, and slow application performance. Each problem includes its potential symptoms for diagnosis.
π Lecture Summary
Broadcast or Multicast Storms
Broadcasts are typically not forwarded so problems are limited to a single network. Multicasts can be forwarded through an internetwork, so they can cause a greater problem if something goes wrong and they storm the network.
π Definition β Broadcast/Multicast Storm: An overwhelming flood of packets addressed to the all-nets broadcast (255.255.255.255) or a multicast address (224-239.x.x.x), causing network degradation.
π‘ Why this matters: Multicasts can propagate across routers, so a single faulty host can overwhelm an entire enterprise network.
Switch Loop
Older switches that have not been restarted or rechecked periodically may malfunction. If there is a loopβnetwork will become overwhelmed with looped packets.
π Definition β Switch Loop: A network condition where packets circulate endlessly between switches, multiplying and consuming bandwidth. π Example: A high rate of identical packets appearing repeatedly on the network indicates a switch loop.
Virus/Malware on Network Hosts
When a compromised host begins performing port scans on other network hosts or begins broadcasting discovery packets, the overhead may be felt.
π Definition β Virus/Malware Overhead: Network congestion caused by infected hosts conducting unauthorized scans or broadcasts. π Example: Local broadcasts appearing frequently, or unusual internal targets (like a local client trying to connect to the Accounting server), indicate malware activity.
Network Name Resolution
This is imperative to connecting to a target. A user may receive an application error such as "Server not found" as in the case of a DNS Name Error when browsing.
π Definition β Name Resolution Failure: The inability to translate a hostname to an IP address, resulting in connection failures. π Example: No response to a name query, or an error response to a name query (such as a DNS Name Error or Server Error).
Network Address Resolution
Incorrect subnet addressing can cause a client to send packets destined to a local device to a router.
π Definition β Subnet Misconfiguration: When a host's IP address or subnet mask is wrong, causing it to treat local devices as remote. π Example: ARP Requests are sent for remote targets. Traffic destined to local hosts is sent to the router instead.
Hardware Address Resolution
This allows obtaining MAC address of a local target or local router. Network address resolution problems may cause a host to think local devices are remote (or vice versa).
π Definition β ARP Resolution Failure: The inability to resolve a local IP address to its corresponding MAC address. π Example: No response to an ARP query for a local host. No response to an ARP query for a local router.
No Support for Selective ACK
When packet loss is detected by a receiving TCP host that supports SACK, that host can acknowledge data received after the missing packet(s). Without SACK, the first packet lost and every subsequent data packet will be retransmitted even when those subsequent packets were received successfully.
π Definition β SACK (Selective Acknowledgment): A TCP option that allows a receiver to acknowledge non-contiguous data, avoiding unnecessary retransmissions. π Example: No SACK Option in the TCP header of a SYN packet. No SACK Option in the TCP header of a SYN/ACK packet when there was a SACK Option in the TCP header of a SYN packet.
No Support for Window Scaling
Window Scaling is used to increase the advertised TCP receive buffer size past the 65,535-byte limit caused by the 2-byte Window Size field. During the TCP handshake, peers indicate that they support Window Scaling and provide a Window Scale Shift Count. Reduces delays caused by a low window or Zero Window condition.
π Definition β Window Scaling: A TCP option that multiplies the Window Size field to allow larger receive buffers, improving throughput on high-latency links. π Example: No Window Scaling Option in TCP header of a SYN packet. No Window Scaling Option in TCP header of a SYN/ACK packet when there was a Window Scaling Option in TCP header of a SYN packet.
Client Misconfiguration
A misconfigured client may have the wrong DNS address, the wrong router address, incorrect port numbers defined in a services file, or other problems.
π Definition β Client Misconfiguration: Incorrect network settings on a client device preventing proper communication. π Example: The client sends traffic to the wrong target. The client receives service refusals or no answer. Numerous other symptoms may appear depending on the misconfiguration.
Low Packet Size/Low MTU Size
The amount of data that can be carried in a frame is limited by MTU, MSS, or even an inefficient application. Performance suffers when sending large files.
π Definition β Low MTU/MSS: A network path or endpoint that advertises an unusually small maximum segment or transmission unit size. π Example: The MSS value defined in the TCP SYN or SYN/ACK packet is illogically small (e.g., less than typical 536 or 1460 bytes).
TCP Port Number Reuse
Port numbers can be reused without any problem if the previous TCP connection is terminated. Otherwise, the new connection is refused.
π Definition β Port Reuse Conflict: Attempting to reuse a TCP port that is still in a TIME_WAIT or established state, causing connection refusal. π Example: TCP RST sent in response to a SYN packet to a server port that is known to be open.
Slow Application
Slow applications can be due to poor or bloated coding, internal errors, or even man-made timers defining the application's performance speed.
π Definition β Slow Application: An application that exhibits large delays in response time due to software inefficiencies or intentional timing. π Example: Large delay in the application response time value observed in packet captures.
β Key Takeaways
The eight additional common network problems cover storms (broadcast/multicast), physical loops, malware, name resolution failures, subnet misconfiguration, ARP failures, TCP optimization absences (SACK and Window Scaling), client misconfiguration, low MTU, port reuse conflicts, and slow applications. For each, diagnosis relies on specific symptoms visible in packet capturesβsuch as high identical packet rates for loops, no SACK/Window Scaling options in handshake packets, RST responses to open ports, or small MSS values. Understanding these symptoms enables rapid identification of network performance issues and misconfigurations. The most critical skills are recognizing when broadcasts/multicasts indicate storms, detecting looped traffic, and identifying missing TCP options that degrade throughput.
π§ Quick Revision Questions
- What symptom indicates a switch loop is occurring on the network?
- What is the difference between broadcast and multicast storms in terms of propagation?
- What TCP header symptom reveals that a host does not support Selective ACK?
- What network misconfiguration causes ARP requests to be sent for remote targets instead of local ones?
- What packet-level evidence suggests a client has a misconfigured DNS or router address?
π Lecture 88 β Using a Troubleshooting Checklist
π Overview: This lecture presents a structured troubleshooting checklist for analyzing captured network traffic in a trace file. It emphasizes a methodical approach to verify trace file integrity, isolate the complaining user's traffic, identify and prioritize delays, and check for throughput and other communication issues for both TCP and UDP-based applications.
ποΈ Topics Covered
This topic describes the usage of a Troubleshooting Checklist for analyzing captured traffic saved in a trace file. The order of the checklist may change depending on the troubleshooting issue, but it covers verifying trace file integrity and basic communications, focusing on the complaining userβs traffic, detecting and prioritizing delays, looking for throughput issues, checking miscellaneous traffic characteristics, and examining connection issues for TCP and UDP-based applications, as well as spotting application errors.
π Lecture Summary
Using a Troubleshooting Checklist
Captured traffic is saved in a trace file. When you open a trace file, use a basic troubleshooting checklist. The order in which you go through the checklist may change depending on the troubleshooting issue.
Verify Trace File Integrity and Basic Communications
Look for ACKed Unseen Segment. Verify traffic from the complaining user's machine is visible. Verify resolution process completion such as DNS/ARP.
Focus on Complaining Userβs Traffic
Filter out unrelated traffic. Export related traffic to a separate trace file.
Detect and Prioritize Delays
Sort and identify high delta times. Measure path latency (Round Trip Time) using delta times in the TCP handshake. This can be achieved by measuring delta from TCP SYN to SYN/ACK. Measure server response time. This can be obtained by measuring from request to response.
π Definition β delta time: The time difference between two captured packets, used to measure delays.
π Formula: Path Latency (RTT) = Delta time between TCP SYN and SYN/ACK packet β Plain-English meaning: The round-trip time to establish a connection is measured by the time gap between sending a connection request (SYN) and receiving an acknowledgment (SYN/ACK).
π Example: To measure path latency to a server, capture the TCP three-way handshake. If the user's machine sends a SYN packet at time 0.000 seconds and receives a SYN/ACK packet at time 0.050 seconds, the delta time is 0.050 seconds. Therefore, the path latency (Round Trip Time) is 50 milliseconds.
Look for Throughput Issues
Click on low throughput points to jump to problem spots in the trace file. Look at traffic characteristics at low throughput points.
Check Miscellaneous Traffic Characteristics
Check packet sizes during file transfer. Check for ICMP messages. Check for IP fragmentation.
TCP-Based Application: Determine TCP Connection Issues/Capabilities
Look for unsuccessful TCP handshakes. Examine the TCP handshakes Options area, e.g. MSS, SACK.
UDP-Based Application: Identify Communication Issues
Look for unsuccessful requests.
Spot Application Errors
Filter for application error response codes.
β Key Takeaways
A structured troubleshooting checklist is essential for efficiently analyzing network trace files. The process begins by verifying trace file integrity and basic communications, followed by isolating the complaining userβs traffic. To identify problems, you must detect and prioritize delays by measuring path latency (using the delta from TCP SYN to SYN/ACK) and server response time, and look for throughput issues by examining low throughput points. Additionally, checking miscellaneous characteristics like packet sizes, ICMP messages, and IP fragmentation, as well as examining TCP connection options (MSS, SACK) and UDP request success, is critical for comprehensive analysis. Finally, directly filtering for application error response codes is a practical method to spot application-layer errors.
π§ Quick Revision Questions
- What is the first step when opening a trace file in the troubleshooting checklist?
- How do you calculate path latency (Round Trip Time) from a TCP handshake capture?
- Besides path latency, what other delay is measured using delta times in the troubleshooting checklist?
- What should you look for to identify problems in a TCP-based application during a troubleshooting analysis?
- What is one of the miscellaneous traffic characteristics that should be checked during a file transfer?
π Lecture 89 β Wireshark Lab 1
π Overview: This lecture introduces the process of creating a custom Troubleshooting Profile in Wireshark. It matters because profiles allow users to customize Wiresharkβs interface with specific buttons, color schemes, and settings for different tasks, enabling efficient switching between use cases like VoIP analysis, WLAN analysis, or general troubleshooting.
ποΈ Topics Covered
The lecture covers the concept of Wireshark profiles, including the Default profile versus custom profiles. It provides a step-by-step guide to creating a new profile named "Troubleshooting Book Profile" using the "Classic" global configuration, which employs the most vibrant colors. It also explains how to access the profile via the Status Bar and the Configuration Profile window.
π Lecture Summary
Creating Your Troubleshooting Profile
Until you create a new profile, you work in Wireshark's Default profile. The profile you are currently working in is shown in the right-side column of the Status Bar. You can create separate profiles for different needs, such as a VoIP profile, a WLAN profile, and a general troubleshooting profile. Creating a custom profile allows you to customize Wireshark with buttons, colors, and more.
π Definition β Profile: A set of customized settings in Wireshark (including buttons, colors, and filters) that can be saved and switched between for different tasks.
π‘ Why this matters: Profiles let you instantly switch between different environments (e.g., troubleshooting vs. VoIP analysis) without manually reconfiguring Wireshark each time.
π Example: Creating the Troubleshooting Book Profile To create a profile from the Classic global configuration:
- Right-click the Profile column on the Status Bar.
- In the Configuration Profile window, select New.
- Click the arrow in the Create from area, expand the Global section, and select Classic (this profile uses the most vibrant colors).
- Enter Troubleshooting Book Profile in the Profile Name area.
- Click OK.
The resulting profile (shown in Figure 89.2) is now ready to use.
β Key Takeaways
The most critical thing to remember is that the Default profile is active until a custom profile is created, and you can see the current profile in the right side of the Status Bar. Custom profiles, such as the Troubleshooting Book Profile, are created via the Configuration Profile window by right-clicking the Status Bar. The Classic global configuration should be selected to get vibrant colors. Finally, multiple profiles can be created and quickly switched between for different troubleshooting needs.
π§ Quick Revision Questions
- Where is the current Wireshark profile displayed?
- What are the four steps to create a custom Troubleshooting profile?
- Which global configuration option is recommended in this lecture and why?
- Name three different types of profiles that might be created for different needs.
- What is the name of the profile created in this lecture?
π Lecture 90 β Wireshark Lab 2
π Overview: This lecture teaches you how to enhance the Packet List Pane Columns in Wireshark to display additional packet information for faster analysis. You will learn to add custom columns, specifically focusing on HTTP delay measurement by extracting the "Time since request" field from HTTP packets.
ποΈ Topics Covered
This lecture covers enhancing the default Packet List Pane columns in Wireshark by adding custom columns. The focus is on adding an HTTP Delta column to measure time delays between HTTP requests and responses. The demonstration uses the tr-httpdelta.pcapng trace file, which contains Windows update and virus detection traffic.
π Lecture Summary
Topic 90: Wireshark Lab 2
This topic discusses how to enhance the Packet List Pane Columns in Wireshark. By default, the Packet List pane contains: No. (number), Time, Source, Destination, Protocol, Length, and Info columns. You can add columns to display additional information about packets to speed up your analysis process.
Step 1: Open tr-httpdelta.pcapng. This trace file contains traffic to/from a user's machine that is checking for Windows updates as well as virus detection updates.
Step 2: Packets 1-3 are TCP handshake packets. Packet 4 is an HTTP GET request for a file called minitri.flg. Packet 5 is an ACK for the GET request. Packet 6 is the HTTP 200 OK response.
Select Packet 6 in the Packet List pane and then, in the Packet Details pane, expand the Hypertext Transfer Protocol.
Step 3: Scroll to the bottom of the HTTP section and right-click on the [Time since request: 0.019036000 seconds] line. Select Apply as Column.
Step 4: Wireshark places the new Time since Request column to the left of the Info column. Right-click on this new column heading and select Edit Column Details. Enter HTTP Delta in the Title area. Click OK.
π‘ Why this matters: By adding the HTTP Delta column, you can now quickly scan and identify HTTP delays between requests and responses without manually calculating them, significantly speeding up network troubleshooting.
β Key Takeaways
The most critical thing to remember is that Wireshark allows you to create custom columns by right-clicking any value in the Packet Details pane and selecting "Apply as Column." This lecture specifically demonstrates creating an HTTP Delta column to measure the time between an HTTP request and its response, which is invaluable for identifying latency issues. The steps are: open the trace file, select the HTTP response packet, expand the HTTP section, right-click the "Time since request" value, and rename the column to "HTTP Delta." This custom column will now appear in the Packet List pane, enabling rapid identification of slow HTTP transactions.
π§ Quick Revision Questions
- What is the default purpose of the Packet List Pane columns in Wireshark, and why would you add custom columns?
- What is the name of the trace file used in this lab, and what type of network traffic does it contain?
- What is the exact name of the field you right-click on to create the HTTP Delta column?
- After applying the column, what menu option do you use to rename it from "Time since Request" to "HTTP Delta"?
- Which packet number is the HTTP 200 OK response that contains the "Time since request" information?
π Lecture 91 β Wireshark Lab 3
π Overview: This lecture teaches how to modify the Time column settings in Wireshark to measure time intervals between displayed packets. The focus is on changing from the default "Seconds Since Beginning of Capture" to "Seconds since Previous Displayed Packet" for precise delta time calculations.
ποΈ Topics Covered
Changing the Time column display format from seconds since beginning of capture to seconds since previous displayed packet; sorting by HTTP delta time to identify delays; hiding and restoring columns in Wireshark; working with the tr-australia.pcapng trace file containing a web browsing session.
π Lecture Summary
[Wireshark Lab 3 β Changing Time Column Settings]
Packets are time stamped at the moment they are captured. By default, Wireshark sets the Time column to Seconds Since Beginning of Capture. Additionally, the resolution is set to nanoseconds, regardless of whether the packet timestamps contain that level of precision.
Step 1: Open tr-australia.pcapng. This trace file contains a web browsing session and was captured at the client. The goal is to change the Time column setting so we can quickly measure the delta time between displayed packets.
Step 2: This trace file begins with a DNS query and response. TCP connection establishment begins in Packet 3. Select View | Time Display Format | Seconds since Previous Displayed Packet.
π‘ Why this matters: Changing to seconds since previous displayed packet allows you to see the exact time gap between consecutive packets in the display, which is essential for detecting network delays and performance bottlenecks.
[Step 3 β Creating an HTTP Delta Column]
Step 3: Select Statistics | HTTP | Packet Counter | Create HTTP Delta Column. Wireshark creates a new column named HTTP Delta with delta time for each HTTP transaction based on the current time format.
[Step 4 β Sorting by Delta Time]
Step 4: Click on the new HTTP Delta column header to sort the column data from high to low.
[Step 5 β Identifying Delays]
Step 5: Click twice on your new HTTP Delta column header to sort the column data from high to low. Wireshark indicates there is a 2.807332 second delay before one of the HTTP 200 OK responses (Packet 49).
π Definition β Delta time: The time difference between two displayed packets, used to measure network delays. π Formula: Delta time = Time of Current Packet - Time of Previous Displayed Packet π Example: In Packet 49, the HTTP 200 OK response shows a delta time of 2.807332 seconds, indicating a significant delay before this response.
[Step 6 β Hiding and Restoring Columns]
Step 6: Right-click on your HTTP Delta column heading and select Hide Column. You can restore this hidden column at any time by right-clicking on any column header, selecting Displayed Columns, and selecting the column to restore.
β Key Takeaways
The critical skill from this lecture is changing the Time column display format from "Seconds Since Beginning of Capture" to "Seconds since Previous Displayed Packet" using View | Time Display Format. The HTTP Delta column is created via Statistics | HTTP | Packet Counter, enabling precise measurement of delays between HTTP transactions. Sorting this column reveals the largest delays, such as the 2.807332 second delay before Packet 49's HTTP 200 OK. Column hiding and restoring is accomplished through right-click context menus and Displayed Columns settings.
π§ Quick Revision Questions
- What is the default Time column display format in Wireshark?
- Which menu path changes the Time column to show seconds since the previous displayed packet?
- How do you create an HTTP Delta column in Wireshark?
- What does a delta time of 2.807332 seconds indicate in an HTTP response packet?
- How can you restore a hidden column in Wireshark?
π Lecture 92 β Wireshark Lab 4
π Overview: This lecture demonstrates how to apply display filters in Wireshark to isolate traffic between specific hosts, subnets, or conversations. It is essential for network troubleshooting, as it allows you to focus on relevant data when a trace file contains traffic from many hosts.
ποΈ Topics Covered
This topic covers filtering traffic in Wireshark by host address, subnet address, or conversation. It shows how to navigate to the Address Resolution table, identify specific hosts, and use display filters to isolate and save a single conversation between a local client and a remote server (cnn.com).
π Lecture Summary
Topic 92: Wireshark Lab 4
This topic describes how to apply a filter on a host, subnet, or conversation in Wireshark. When you capture traffic at the server or inside the network infrastructure, your trace file may contain conversations between many hosts on the network. Are you interested in the traffic between a specific client and server? If yes, you can apply a display filter based on a host address, a subnet address, or a conversation.
Step 1: Open tr-cnn.pcapng.
Step 2: This trace file contains numerous conversations. Letβs extract conversations between the local client and cnn.com servers. We begin with looking at the name resolution information that Wireshark extracted from the trace file. Select Statistics | Show Address Resolution.
[Figure 92.1: Extract and Save a Single Conversation]
β Key Takeaways
A student must remember that a captured trace file often contains traffic from many hosts, so filtering is critical for focused analysis. The key method is to use Statistics > Show Address Resolution to view the resolved hostnames. From this table, you can identify the specific IP addresses for a client-server pair and then build a display filter to isolate their conversation. This technique is essential for extracting and saving only the relevant packets from a large capture.
π§ Quick Revision Questions
- What is the first step to isolate traffic between a specific client and server in a Wireshark trace?
- Which Wireshark menu path is used to view the resolved hostnames from the trace file?
- When capturing traffic inside network infrastructure, why is it necessary to apply a display filter?
- What is the purpose of the "Show Address Resolution" feature in Wireshark?
- How would you filter traffic to see only conversations between a local client and cnn.com servers?
π Lecture 93 β Wireshark Lab 5
π Overview: This lecture focuses on advanced Wireshark filtering techniques, specifically how to apply display filters using port numbers and application names to isolate specific network conversations. These skills are essential for analyzing network traffic efficiently, especially when dealing with large trace files containing thousands of packets.
ποΈ Topics Covered
The lecture covers how to filter traffic from a specific domain (cnn.com) using IP address ranges, how to export filtered packets to separate trace files, how to apply conversation filters via right-click, and how to filter based on port numbers versus application names such as FTP data. It includes step-by-step Wireshark instructions for each technique.
π Lecture Summary
Filtering Traffic from cnn.com Domain
All servers in the cnn.com domain begin with the IP prefix 157.166. In the display filter area, enter ip.addr==157.166.0.0/16 and click Apply. The Status Bar indicates that 360 packets match this filter, listing all conversations to and from the cnn.com servers. Use File | Export Specified Packets (with the Displayed radio button selected by default) and name your file tr-cnntraffic.pcapng to save this filtered traffic.
To filter a single conversation, use the right-click method on a packet in the Packet List pane. Select Conversation Filter | TCP. For example, on Packet 3, this results in 55 packets appearing. Export this conversation by selecting File | Export Specified Packets and naming the file tr-cnnconv1.pcapng. Always click the Clear button to remove your display filter afterwardβavoiding unrelated traffic and potential distractions makes analysis easier. Therefore, save the traffic from an interesting conversation to a separate trace file.
Wireshark Lab 5: Filtering by Port Number
There are two ways to define a display filter on an application in a trace file: filter based on the application name (if known to Wireshark) or the port number in use. Open tr-twohosts.pcapng and look at the FTP data transfer connection established by 192.168.1.119.
Enter ip.addr==192.168.1.119 in the display filter area and click Apply. This filters all traffic to and from 192.168.1.119. Look for the response to the PASV command (Packet 3,959). Expand the FTP section in the Packet Details pane to see the port number that the server will be listening on for the FTP data channel (port 39,757).
π Definition β PASV command: Passive mode FTP command where the server opens a port and waits for the client to connect for data transfer.
Now replace your address display filter with tcp.port==39757 and click Apply. The Status Bar indicates that 28,020 packets match this filter. Next, contrast this with a display filter based on an application name.
Replace your TCP port filter with ftp-data and click Apply. Now you wonβt see the TCP handshake, ACKs, or the connection teardown packetsβthe filter only shows packets actually carrying FTP data payload.
π Example: When filtering by tcp.port==39757, you see 28,020 packets including the TCP three-way handshake, acknowledgments, and connection close. When filtering by ftp-data, you see only the packets containing actual file transfer data, not the protocol overhead.
Click the Clear button to remove your display filter.
β Key Takeaways
A student must remember that Wireshark allows two primary methods for filtering application traffic: by port number (e.g., tcp.port==39757) and by application name (e.g., ftp-data), with the port filter capturing all protocol overhead packets while the application name filter shows only data-bearing packets. The right-click conversation filter quickly isolates a single TCP conversation. Saving filtered traffic to separate trace files (using Export Specified Packets) keeps analysis organized and distraction-free. Always clear display filters after use to avoid unintentional filtering.
π§ Quick Revision Questions
- What IP address prefix do all cnn.com servers use, and how do you filter for that entire network in Wireshark?
- After filtering traffic from 192.168.1.119, where do you find the port number that the FTP server will use for the data channel?
- What is the difference in the packets shown when using
tcp.port==39757versusftp-dataas a display filter? - How do you save a specific conversation from a display filter to a separate trace file?
- Why should you click the Clear button after applying and saving a display filter?
π Lecture 95 β Wireshark Lab 7
π Overview: This lecture describes how to locate buffer problems using the Calculated Window Size field in Wireshark. It demonstrates how to identify Zero Window conditions and low window size values that can stop TCP data transmission.
ποΈ Topics Covered
This topic covers the significance of the TCP Window Size field in detecting buffer problems, how to apply a filter to identify packets with low window size values, and the impact of insufficient buffer space on data transmission between TCP peers.
π Lecture Summary
Locating Buffer Problems using the Calculated Window Size Field
Every TCP packet sent by a host contains that host's available receive buffer space in the Window Size field. If Window Scaling is in use, this field value is multiplied by a scaling factor. When the advertised buffer space becomes equal to zero, the host cannot accept any more dataβa Zero Window condition. Even a low Window Size value can stop a TCP peer from transmitting data.
π‘ Why this matters: Identifying low window sizes is critical for troubleshooting network performance issues where data transmission stalls unexpectedly.
Step 1: Open tr-winsize.pcapng.
Step 2: Expand any TCP header in the Packet Details pane. Right-click on the Calculated window size field and select Prepare a Filter | Selected.
Step 3: Change the display filter value to tcp.window_size < 1000 and click Apply. This displays Packet 374 in which the client is advertising a 536-byte receive buffer.
What will happen if the TCP peer has more than 536 bytes of data queued to transmit? This low Window size value will stop data transmission.
π Definition β Zero Window: A condition where a host's advertised buffer space becomes equal to zero, meaning the host cannot accept any more data from its TCP peer.
π Formula: Effective Window Size = Window Size field Γ Scaling Factor (when Window Scaling is active)
π Example: In the trace file tr-winsize.pcapng, Packet 374 shows a client advertising a 536-byte receive buffer. The server had more than 536 bytes queued to transmit but could not send a full-sized packet because the client only had 536 bytes of buffer space available. The server had to wait until the client's buffer size increased.
Step 4: Click Clear to remove your filter. In this trace file, we next view the delay that occurs before the Window Update packet (Packet 375).
The Window Update packet is sent when the client's buffer space becomes available again, signaling the server to resume transmission.
β Key Takeaways
The TCP Window Size field in every packet contains the host's available receive buffer space, and a low value can stop data transmission. A Zero Window condition occurs when buffer space reaches zero, preventing any data acceptance. The calculated window size equals the window size field multiplied by the scaling factor when Window Scaling is enabled. Using the filter tcp.window_size < 1000 in Wireshark helps locate buffer problems, and the Window Update packet signals when buffer space becomes available again, allowing transmission to resume.
π§ Quick Revision Questions
- What does the TCP Window Size field in a packet represent?
- What is a Zero Window condition and why does it stop data transmission?
- What Wireshark display filter would you use to find packets where the client's receive buffer is less than 1000 bytes?
- What is the purpose of a Window Update packet?
- What happens when Window Scaling is in use and how does it affect the advertised buffer space?
π Lecture 96 β Wireshark Lab 8
π Overview: This lecture teaches how to filter out "normal" network traffic in Wireshark to focus on anomalies. It covers two methods for excluding traffic based on field values and explains when to use each method correctly. This skill is critical for network forensics and troubleshooting.
ποΈ Topics Covered
The lecture covers filtering out application traffic using the exclamation sign (!), two methods for excluding traffic based on field values (! with == vs. !=), guidelines for choosing the correct method, examples of correct and incorrect display filters, and a step-by-step lab exercise using tr-general.pcapng to reveal hidden Dropbox traffic.
π Lecture Summary
Filtering OUT Traffic
To filter out traffic based on an application name, precede the application display filter name with an exclamation sign (!). For example, to remove ARP from view, use !arp. There are two methods to exclude traffic based on a field value. One method uses the ! or not operator with == or eq. The other uses the != or ne operator.
π Definition β Display Filter: A Wireshark filter that controls which packets are displayed, without affecting packet capture.
π Example β Excluding a specific IP address:
β’ !ip.addr==10.10.10.10 (method 1: ! with ==)
β’ http.request.method != "GET" (method 2: !=)
π‘ Why this matters: Using the wrong operator with multi-field names like ip.addr can produce incorrect results, making you miss anomalies.
When to Use ! and == and When to Use !=
Use the first method (! and ==) when you filter on a field name that matches two fields such as ip.addr, tcp.port, or udp.port. Use the second method (!=) when you refer to a field name that only matches one field such as dns.flags.rcode or tcp.dstport.
π Rule: For multi-field names β !fieldname==value. For single-field names β fieldname!=value.
π Example β Correct vs. Incorrect filters:
β’ Correct: !ip.addr==10.1.1.1
β’ Incorrect: ip.addr != 10.1.1.1 (because ip.addr matches both source and destination)
β’ Correct: dns.flags.rcode!=0
β’ Incorrect: !dns.flags.rcode==0 (because dns.flags.rcode matches only one field)
Lab Exercise: Removing Normal Traffic
This lab removes a set of applications and protocols from view to determine what other traffic is seen on the network.
π Step 1: Open tr-general.pcapng.
π Step 2: In the display filter area, type !tcp && !arp and click Apply. There are 40 packets that match this filter.
π Step 3: To remove DNS and DHCP from view, expand the filter by adding && !dns && !bootp. Eight packets are displayed.
π Example β The displayed packets indicate that there are two hosts running Dropbox on the network and are sending packets to the broadcast address (255.255.255.255).
β Key Takeaways
The most critical exam point is the correct syntax for excluding traffic: use !field==value for multi-field names (ip.addr, tcp.port, udp.port) and field!=value for single-field names (dns.flags.rcode, tcp.dstport). The lab demonstrates a systematic approach: start with the most common protocols (TCP, ARP), then add DNS and DHCP to reveal hidden application traffic like Dropbox broadcasts. Remember that ip.addr != 10.1.1.1 is incorrect because ip.addr matches both source and destination IP fields, while !ip.addr==10.1.1.1 works correctly. The exclamation sign (!) is the universal operator for excluding protocols by name (e.g., !arp, !dns, !bootp).
π§ Quick Revision Questions
- What is the correct display filter to exclude all traffic from IP address 10.1.1.1?
- Why is
ip.addr != 10.1.1.1an incorrect filter while!ip.addr==10.1.1.1is correct? - How many packets remain after applying
!tcp && !arpto tr-general.pcapng? - What operators are added to remove DNS and DHCP traffic in the lab?
- What two hosts were revealed after removing normal traffic in the lab, and what type of packets were they sending?
π Lecture 97 β Wireshark Lab 9
π Overview: This lecture explains how to create custom Filter Expression Buttons in Wireshark to quickly apply display filters for identifying common network problems. The specific example demonstrates building a button that detects TCP handshake packets missing Selective Acknowledgment (SACK) and Window Scaling functionality.
ποΈ Topics Covered
Creating Filter Expression Buttons in Wireshark using display filters; building a combined inclusion filter for SYN bits and exclusion filters for TCP options; applying the filter to the sample trace file tr-smbjoindomain.pcapng; saving the filter as a reusable button named TCP-HS.
π Lecture Summary
Creating Filter Expression Buttons
Filter Expression buttons are based on display filters. These buttons can be created and used to quickly apply display filters to your traffic to identify common network problems. The lecture demonstrates creating a button to quickly identify TCP handshake packets that do not offer Selective Acknowledgment (SACK) or Window Scaling functionality. To do this, we need to combine an inclusion filter for the SYN bit set to 1 with an exclusion filter for the SACK and Window Scaling options.
π‘ Why this matters: Instead of manually typing complex filters each time, you can create one-click buttons that instantly highlight problematic packets, saving significant time during network troubleshooting.
Step-by-Step Procedure
Step 1: Open tr-smbjoindomain.pcapng.
Step 2: Packet 11 is the first SYN packet in the trace file. Right-click on the TCP header of this packet and select Expand Subtrees. First build the filter and then turn the filter into a filter expression button. Right-click on SYN: Set line and select Prepare a Filter | Selected. Wireshark places the first part of the filter in the display filter area.
![Figure 97.1: Create a Button to Detect Missing TCP Functionality]
Step 3: Scroll down to the TCP Options area. Click on the TCP SACK Permitted Option: True line. This displays tcp.options.sack_perm in the Status Bar area. Expand the filter by typing && !tcp.options.sack_perm so that TCP handshake packets that do not contain this value can be observed.
π Definition β tcp.options.sack_perm: A TCP option field that indicates the SACK Permitted option is present in the TCP handshake.
π Filter logic: tcp.flags.syn == 1 && !tcp.options.sack_perm β Shows SYN packets that do NOT have SACK permitted enabled.
Step 4: Now add || ! tcp.options.wscale.multiplier and put parentheses around the options, as shown next.
π Definition β tcp.options.wscale.multiplier: A TCP option field containing the Window Scale multiplier value.
π Complete filter: tcp.flags.syn == 1 && (!tcp.options.sack_perm || !tcp.options.wscale.multiplier) β Shows SYN packets that lack EITHER SACK OR Window Scaling (or both).
Step 5: Click the Save button and name your new button TCP-HS and click OK.
β Key Takeaways
Filter Expression Buttons provide one-click access to complex display filters for rapid network problem identification. The lecture demonstrates building a combined filter that detects TCP handshake packets missing SACK or Window Scaling functionality, both critical TCP optimizations for modern high-performance networks. The filter uses tcp.flags.syn == 1 to isolate handshake packets, then excludes those with tcp.options.sack_perm or tcp.options.wscale.multiplier. Parentheses ensure correct logical grouping when combining multiple exclusion conditions. The saved button (TCP-HS) can be reused across any capture file for instant TCP handshake analysis.
π§ Quick Revision Questions
- What two TCP options are being checked for absence in the filter expression button created in this lab?
- Why must parentheses be placed around the option exclusion part of the filter?
- Which Wireshark menu option is used to begin building the display filter from the SYN field?
- What does the filter condition
!tcp.options.sack_permspecifically exclude from the display? - What is the purpose of using a Filter Expression Button instead of typing the filter each time?