Compare Check Point fw monitor and tcpdump, choose capture points, and collect a focused R82 packet trace for connectivity issues.
Choose the evidence you need
Use an interface capture to establish whether a request or response is present at a selected network boundary. Use fw monitor when you need to investigate packet handling at Check Point inspection points. Their views are complementary. The precise visibility depends on software release, acceleration, virtual system, and capture configuration.
Scope and prerequisites
The fw monitor syntax below is based on the R82 CLI reference. Run authorized diagnostics on the gateway carrying the traffic and use the appropriate Expert shell privileges. Record the active cluster member and virtual system. Confirm available disk space before writing a capture, and use a unique file name so that an earlier capture is preserved.
Start with a bounded interface capture
tcpdump -ni eth1 -s 0 -c 100 'host 10.10.10.25 and tcp port 443'
What the filter establishes
Replace eth1 with the actual interface. The filter captures up to 100 matching packets involving the example client and TCP port 443 on that interface. It includes either direction if the matching packets traverse it. The -s 0 option requests the available packet content, which can expose application data. Adjust the capture length to the incident’s evidence requirements.
Use a five-tuple fw monitor filter
fw monitor -F "10.10.10.25,0,10.40.20.50,443,6" -F "10.40.20.50,443,10.10.10.25,0,6" -o /var/log/fsc-https-case01.cap
Read the filter before running it
In this R82 example, the five positions represent source IP, source port, destination IP, destination port, and protocol number. Zero is the wildcard used for the unspecified port. The second filter covers the reverse tuple. NAT can change the addresses visible at later points, so add the relevant translated tuple when the investigation requires it. Stop the capture with Ctrl+C after reproducing the problem.
Correlate the two views
| Question | Useful evidence |
|---|---|
| Does the client request reach the gateway? | Capture on the actual ingress interface |
| What changes during handling? | fw monitor observations at the relevant inspection points |
| Does a translated request leave? | Egress capture matched against NAT expectations |
| Does the server response return? | Capture on the server-facing path and inspection evidence |
Avoid an incomplete conclusion
A packet missing from a chosen point requires a check of the filter, member, interface, virtual system, and acceleration behavior. Interpret i, I, o, and O observations using the documentation for the installed release. Do not assume every packet must appear at every point under every architecture. An interface trace alone also cannot prove which security rule handled the packet.
Validation and handover
Stop both captures, preserve their timestamps, and test the application again after the selected repair. Include the original and translated tuples, software version, capture locations, and relevant SmartConsole logs in the incident record. Protect captured files according to their content and retention requirements.