04 Networking · TShark · Troubleshooting
Four packet captures, four network faults
Four short cases cover DNS failure, a TCP reset, duplicate-IP ARP, and path MTU, each with a post-change capture.
- Issue
- Which small set of frames explains the symptom, and what separate capture shows that the condition changed?
- Decisive evidence
- Each case links the user’s symptom to the relevant packet fields and a new capture taken after the change.E01
- Result
- Each note identifies the frames that explain the symptom and keeps the post-change capture separate.
- Boundary
- These are controlled packet cases in an isolated ARM64 lab.Full limit
- Completion
- Complete within stated scope
- Execution
- Isolated ARM64 packet lab
- Review
- Field extraction and manifest review
- Evidence
- Selected evidence on this page
- TShark
- tcpdump
- Capinfos
- Python
Project scope
What was built and checked.
Project record: The synthetic cases, protocol fields that explain each symptom, failure and post-change captures, and the support interpretation.
Method
Start with the symptom, apply one relevant filter, then capture the post-change result.
- 01
State where the capture was taken and what the user observed.
- 02
Select only the protocol fields needed to test the suspected fault.
- 03
Record the exact change outside the capture.
- 04
Use a separate post-change capture to confirm the new result.
Four notes
Short records for four different symptoms.
DNS name failure
- Symptom
- The service could not be opened by name.
- Evidence
- The DNS response returned NXDOMAIN for the requested record.
- Post-change
- The post-change capture returned the expected A record.
- Limit
- This note does not identify why the DNS record was wrong.
TCP reset
- Symptom
- The client reached the host, but the TCP connection was refused immediately.
- Evidence
- The service address returned a TCP reset in response to the connection attempt.
- Post-change
- After the listener started, the separate post-change capture completed the handshake and carried the test payload.
- Limit
- The capture proves transport behavior, not the application configuration change.
Duplicate-IP ARP
- Symptom
- Connectivity changed between two devices using the same address.
- Evidence
- ARP replies associated one IPv4 address with two different MAC addresses.
- Post-change
- The post-change capture showed one remaining owner.
- Limit
- The case uses synthetic devices and does not claim switch-level remediation.
Path MTU
- Symptom
- Small traffic worked while a larger payload failed.
- Evidence
- ICMP type 3, code 4 reported that fragmentation was needed for the path.
- Post-change
- The post-change capture delivered the same test payload without that response.
- Limit
- The capture identifies the path symptom, not every possible MTU cause.
Evidence from the lab
The checks used to reach the conclusion.
Selected outputs from the lab. Sensitive and unnecessary details are omitted.
Failure : rcode 3, NXDOMAIN
Post-change : rcode 0, A 198.51.100.11Failure : reset frame present
Post-change : zero resets, echo verifiedFailure owners : 2
Post-change owners : 1Full evidence archive2 additional artifacts+
Failure : ICMP type 3, code 4, MTU 1200
Post-change : same DF payload deliveredTShark / Capinfos : Wireshark 4.6.7
Arguments : sanitized
Exit status : 0Result and limits
What changed, and what this lab does not prove.
Result
The packet fields distinguish name-resolution failure, a refused TCP connection, duplicate-IP ownership, and path-MTU failure. Each case includes a separate post-change capture.
Limit
These are controlled packet cases in an isolated ARM64 lab. Public downloads are sanitized validation records derived from private captures. They do not claim production monitoring or broad incident response.
Claims and evidence