Overview
This lecture covers interface monitoring for network devices, common ethernet errors, and administrative responses to interface issues.
Interface Monitoring & Automation
- Network administrators monitor interfaces to detect problems like bad cables or congestion before outages occur.
- Interface statistics can reveal network overutilization and guide design adjustments.
- Manual monitoring is inefficient; automation uses Simple Network Management Protocol (SNMP).
- SNMP devices support a Management Information Base (MIB II) for standard statistics; proprietary MIBs provide device-specific data.
Key Interface Statistics
- Link status determines if a connection is up (working) or down (not working).
- Utilization metrics track bandwidth and throughput to ensure adequate service performance.
- Monitoring errors like CRC errors, runts, giants, and drops can preempt network failures.
- Bandwidth tests help assess actual throughput over a connection.
Ethernet Frame Structure & Errors
- Ethernet frames start with a Preamble and Start Frame Delimiter (SFD), followed by destination/source MAC addresses, EtherType, payload, and Frame Check Sequence (FCS).
- FCS uses a Cyclic Redundancy Check (CRC) to detect corruption; CRC errors indicate data problems, often due to bad cables or interfaces.
- Runts are frames smaller than 64 bytes, usually caused by collisions in half-duplex networks.
- Giants are frames larger than the standard maximum (1,518 bytes); jumbo frames have a higher configured maximum.
- Drops occur when frames cannot be buffered and are lost.
Error Handling and Administrative Actions
- Most operating systems allow viewing of interface error counters for diagnostics.
- Incrementing CRC errors often mean hardware issues needing repair or replacement.
- Administrators can manually disable problematic interfaces to prevent wider network impacts.
- Devices can automatically error-disable interfaces when persistent issues (flapping, security violations, excessive errors) are detected.
- Error-disabled interfaces must be manually re-enabled by an administrator.
- Administratively down interfaces are intentionally disabled by an administrator.
- Suspended port status occurs when configuration mismatches (e.g., incorrect link aggregation settings) are detected on activation.
Key Terms & Definitions
- SNMP — Simple Network Management Protocol, used for automated network monitoring.
- MIB — Management Information Base, a database of monitored network device statistics.
- CRC Error — Cyclic Redundancy Check error, indicates corrupted data frames.
- Runts — Ethernet frames under 64 bytes, generally error frames.
- Giants — Ethernet frames exceeding the standard maximum size.
- Drop — Lost frames due to buffering or congestion issues.
- Error Disabled — Interface automatically shut down due to persistent errors.
- Administratively Down — Interface manually disabled by a network administrator.
- Suspended — Interface disabled due to configuration mismatch at activation.
Action Items / Next Steps
- Review error counters on network devices for signs of hardware or configuration issues.
- Study Ethernet frame structure and common error types for upcoming assessments.