Description

This article explains the below log message when received on MX device.

 

ASIC Error detected errorno 0x000700fa

Symptoms

An FPC might get rebooted causing all the links to go down and a Major alarm reported. Below logs will be observed.

 

May 9 20:43:53 lab-router : %PFE-3: fpc2 XMCHIP(0): XMCHIP(0): EPM0: Protect: Log Error 0x1, Log Address 0x299, Multiple Errors 0x0 

May 9 20:43:53 lab-router chassisd[8866]: %DAEMON-3-CHASSISD_FPC_ASIC_ERROR: <FPC 2> ASIC Error detected errorno 0x000700fa (null)

May 9 20:43:53 lab-router alarmd[9040]: %DAEMON-4: Alarm set: FPC id=167772776, color=YELLOW, class=CHASSIS, reason=FPC 2 Minor Errors

May 9 20:43:53 lab-router craftd[8870]: %DAEMON-4: Minor alarm set, FPC 2 Minor Errors

May 9 20:43:55 lab-router : %PFE-3: fpc2 Cmerror Op Set: XMCHIP(0): XMCHIP(0): EPM0: Protect: Parity error for free pool single port SRAM (URI: /fpc/2/pfe/0/cm/0/XMCHIP(0)/0/XMCHIP_CMERROR_EPM_PAR_PROTECT_FSET_REG_DETECTED_FP) 

May 9 20:46:53 lab-router rpd[8999]: %DAEMON-4-JTASK_IO_CONNECT_FAILED: AGENTD I/O.128.0.0.4+9500: Connecting to 128.0.0.4+9500 failed: Connection refused

May 9 20:47:47 lab-router : %PFE-3: fpc2 XMCHIP(0): XMCHIP(0): EPM0: Packet delimiter violation detected - Queue 64 

May 9 20:48:00 lab-router alarmd[9040]: %DAEMON-4: Alarm set: FPC id=150995560, color=RED, class=CHASSIS, reason=FPC 2 Major Errors

May 9 20:48:00 lab-router craftd[8870]: %DAEMON-4: Major alarm set, FPC 2 Major Errors

 

Solution

This message is due to a transient hardware memory parity error in the EPM block on the XMCHIP ASIC. This error is caused by a transient hardware issue. A single occurrence of the message in syslog, not seen in combination with other events, indicates a one-time transient hardware problem. However, multiple continuous occurrences indicate persistent underlying issues.

 

XM is memory chip. EPM is the Egress Priority-Flow-Control Manager Block. EPM0 and EPM1 are for the WAN interfaces (interfaces facing the input/output external devices). Each EPM block serves a subset of the interfaces of the PFE and is responsible for priority-based flow control of packets leaving the ASIC.

 

On the affected FPC, the logs as shown in the Symptom section represent a transient memory issue on the XM chip's EPM block (Egress Priority-Flow-Control Manager).

 

A single occurance should be cleared on its own and no further action is needed however if the logs are coming frequently reboot the affected FPC and that should clear the underlying issue.

 

If the logs are observed after reboot then its a Hardware issue so proceed with the replacement.

 

Modification History

2024-05-24 : Article Created