Edit

Troubleshoot packet drops, discards, and errors for Azure Operator Nexus

Dropped Packets 

This section provides guidance to identify, analyze, and remediate packet drops, discards, and interface‑level errors that may impact network reliability and performance.

What to Check 

  • Validate whether packet drops are expected due to normal forwarding behavior (for example, buffer exhaustion under load). 
  • Review traffic policing and rate‑limiting configuration that may intentionally drop packets. 
  • Correlate drops with traffic spikes or congestion events. 
  • Tune traffic policing or rate‑limit configuration if drops are unintended. 
  • If drops persist and impact traffic, engage Arista Technical Assistance Center (TAC) for further analysis. 

Interface In Discards 

What to Check 

  • Determine whether discards are caused by buffer overflows, congestion, or traffic policing. 
  • Review interface queue health and identify queues experiencing drops. 
  • Identify interfaces reporting non‑zero discard counters and isolate affected ports. 
  • Verify whether Quality of Service (QoS) policies are enforcing drops. 
  • Inspect hardware‑level discard reasons (for example VLAN or routing‑related drops). 
  • Adjust queue configuration or traffic patterns to reduce congestion. 
  • Review and correct QoS or policing policies if misconfigured. 
  • If discards can't be explained or resolved, engage Arista TAC and provide interface counters, queue statistics, and relevant logs. 

Interface In Errors 

What to Check 

  • Identify physical‑layer issues such as cabling faults, optical issues, or transient link instability. 
  • Validate whether malformed or corrupted packets are being received. 
  • Review interface error counters to determine error patterns. 
  • Inspect and replace faulty cables, optics, or transceivers. 
  • Verify link configuration consistency across both ends of the interface. 
  • If errors persist, escalate to Arista TAC for deeper diagnostics. 

Interface Out Discards 

What to Check 

  • Determine whether outbound congestion is causing packet drops. 
  • Review egress queue utilization and buffer behavior. 
  • Check for traffic policing or shaping on outbound interfaces. 
  • Adjust queue allocation or traffic shaping to reduce congestion. 
  • Review QoS policies applied to outbound traffic. 
  • Engage Arista TAC if drops persist or can't be correlated to expected traffic behavior. 

Interface In or Out Errors 

What to Check 

  • Confirm Ethernet frame errors at the data‑link (L2) layer. 
  • Validate speed, duplex, and autonegotiation settings on both ends of the link. 
  • Review optical power levels and physical media health. 
  • Replace faulty cables or transceivers and clean fiber connectors. 
  • Correct speed, duplex, or negotiation mismatches. 
  • If error rates remain high, escalate to Arista TAC with interface diagnostics and optical readings. 

LACP Errors 

What to Check 

  • Validate LACP configuration consistency (mode, timers, LAG identifiers). 
  • Confirm all LAG member ports have identical attributes (speed, MTU, VLAN configuration). 
  • Inspect physical link health for all LAG members. 
  • Verify LACP control traffic isn't blocked by policies or control‑plane features. 
  • Ensure no conflicting LACP controllers or bonding configurations exist. 
  • Correct configuration mismatches across LAG members. 
  • Replace faulty cables, optics, or ports. 
  • Resolve control‑plane or policy conditions blocking LACP PDUs. 
  • Engage Arista TAC if LACP instability persists. 

Ethernet CRC Errors 

What to Check 

  • Identify corrupted frames indicating physical‑layer issues. 
  • Review interface counters on both ends of the link. 
  • Validate optical power levels are within supported ranges. 
  • Replace damaged cables or faulty transceivers. 
  • Clean fiber connectors and reseat optics. 
  • Correct speed, duplex, or negotiation mismatches. 
  • Escalate to Arista TAC if CRC errors continue after remediation.
  • Last updated on