This article addresses syslog messages related to PPE (Packet Processing Engine) traps observed during IPFIX flow record export, specifically when the system fails to retrieve the correct Protocol Next Hop (PNH) for IPv6 flows. The issue is linked to the absence of the ipv6-extended-attrib knob in the configuration.

• During flow export operations, the system may encounter PPE related syslog messages when it fails to resolve the correct PNH, particularly for IPv6 traffic. 
• These messages are typically non-service-impacting but indicate incomplete or incorrect flow record generation, which can affect traffic visibility and analytics.
• Upon reviewing the RSI and var logs output, it was found that the ipv6-extended-attrib knob was not configured on some FPCs.
• This knob is critical for enabling accurate PNH resolution for IPv6 flows.
• Without it, the system may generate incomplete flow records, leading to internal PPE-related logs.
  
  
  

These logs have been observed across multiple FPCs ---

• listmgr_event_handler: Logs events related to flow list management, often triggered during flow export operations.
• xss_event_handler: Handles events related to cross-subsystem signalling, typically involved in coordination between different ASIC blocks.
• ppe_error_interrupt: Indicates error conditions or interrupts raised by the Packet Processing Engine (PPE), often due to unexpected conditions during flow handling or export.

This reinitialization is expected and ensures accurate flow export and suppression of the related syslog messages.

Jul 30 12:30:01.012 fpc5 listmgr_event_handler(856): EA[1:0].llm ILLEGAL_ADD_ERR. 
Jul 30 12:30:01.012 fpc5 listmgr_event_handler(866): EA[1:0].llm ILLEGAL_XTXN. 
Jul 30 12:30:01.012 fpc5 xss_event_handler(1109): EA[1:0]_PPE 48.xss[0] ADDR Error. 
Jul 30 12:30:01.012 fpc5 ppe_error_interrupt(5486): EA[1:0]_PPE 48 Errors sync xtxn error 
Jul 30 12:30:01.012 fpc5 xss_event_handler(1109): EA[1:0]_PPE 49.xss[0] ADDR Error. 
Jul 30 12:30:01.012 fpc5 ppe_error_interrupt(5486): EA[1:0]_PPE 49 Errors sync xtxn error 
Jul 30 12:30:02.044 fpc5 listmgr_event_handler(856): EA[1:0].llm ILLEGAL_ADD_ERR. 
Jul 30 12:30:02.044 fpc5 listmgr_event_handler(866): EA[1:0].llm ILLEGAL_XTXN. 
Jul 30 12:30:02.045 fpc5 xss_event_handler(1109): EA[1:0]_PPE 48.xss[0] ADDR Error. 
Jul 30 12:30:02.045 fpc5 ppe_error_interrupt(5486): EA[1:0]_PPE 48 Errors sync xtxn error 
Jul 30 12:30:02.045 fpc5 xss_event_handler(1109): EA[1:0]_PPE 49.xss[0] ADDR Error. 
Jul 30 12:30:02.045 fpc5 ppe_error_interrupt(5486): EA[1:0]_PPE 49 Errors sync xtxn error 

• > These logs are related to internal flow and packet processing mechanisms within the FPC and are non-service-impacting.
• > They are commonly seen when the system is unable to fetch the correct Protocol Next Hop (PNH) during IPFIX flow record export, especially for IPv6 traffic.
• > Based on the RSI review, it was found that the ipv6-extended-attrib The knob is not configured.
• > This knob is essential for enabling accurate PNH resolution for IPv6 flows.
• > Without it, the system may generate incomplete or incorrect flow records, leading to these internal logs.

Action Required:
Please enable the ipv6-extended-attrib knob to resolve this issue.

Expected Behaviour After Enabling:

• IPFIX sampling will be temporarily disabled.
• All learned flows will be cleared from the flow table.
• The FPC will automatically re-enable IPFIX with the updated configuration.
• Flow learning will resume with proper IPv6 attribute support.

This configuration change will allow the system to correctly interpret extended IPv6 attributes and avoid PPE trap-related syslog messages.

From RSI, we see the following is the config from FPC5 & FPC6 -

sampling-instance ipfix_instance;

For FPC7, we see the following config : 

sampling-instance ipfix_instance;

    inline-services {

      flex-flow-sizing;

      use-extended-flow-memory;

      flow-table-size {

        ipv6-extended-attrib;

      }

    }