This article explains the meaning of the "LKUP_ASIC_PPE_LMEM_ERR (0x40038)" alarm that is seen on MX Series routers and indicates if any actions need to be taken.

MX router equipped with MPC1E/MPC2E/MPC3E/MPC4E/MPC5E/MPC6E cards reports FPC Minor Error alarm.

The example below is shown for MPC3 model of linecard:

user@device> show system alarms

1 alarms currently active

Alarm time               Class  Description

<date> <time>  Minor  FPC 10 Minor Errors

In Var/log Messages:

Mar 8 06:54:24 TEST-MX-RE0 : %PFE-3: fpc10 LMEM errors require LUCHIP(4) PPE 6 Zone 30 disable.

Mar 8 06:54:24 TEST-MX-RE0 : %PFE-5: fpc10 Error: /fpc/10/pfe/0/cm/0/LUCHIP(12)/0/LKUP_ASIC_PPE_LMEM_ERR (0x40038), scope: board, category: functional, severity: major, module: LUCHIP(12), type: PPE LMEM error

Mar 8 06:54:25 TEST-MX-RE0 : %PFE-5: fpc10 Performing action get-state for error /fpc/10/pfe/0/cm/0/LUCHIP(12)/0/LKUP_ASIC_PPE_LMEM_ERR (0x40038) in module: LUCHIP(12) with scope: board category: functional level: major

Mar 8 06:54:35 TEST-MX-RE0 : %PFE-5: fpc10 Performing action disable-pfe for error /fpc/10/pfe/0/cm/0/LUCHIP(12)/0/LKUP_ASIC_PPE_LMEM_ERR (0x40038) in module: LUCHIP(12) with scope: board category: functional level: major

Mar 8 06:54:36 TEST-MX-RE0 : %PFE-3: fpc10 Cmerror Op Set: LUCHIP(12): LMEM errors require LUCHIP(4) PPE 6 Zone 30 disable.

Mar 8 06:54:36 TEST-MX-RE0 : %PFE-6: fpc10 IFFPC: IFD(et-10/0/0, 431) ASIC error notification

In Chassisd:

Jun 25 11:08:16 CHASSISD_FPC_ASIC_ERROR: <FPC 4> ASIC Error detected errorno 0x00040038 (null)

Jun 25 11:08:16 fpc_mx_alarm_request: chasd.support_pic_alarm_from_pic_fru=0 

This error is caused by a transient hardware error. LMEM data errors are parity errors and the software will repair those errors automatically. If the number of LMEM data errors exceeds the threshold of 6, then a particular zone will be disabled from further usage. If this threshold is reached for shared zones, the entire Packet Processing Engine (PPE) will become disabled.

To summarize:

• LMEM data error is a transient hardware problem that gets automatically corrected.
• An alarm is raised if specific zones have many repairs, then the specific zone is disabled.
• If the specific zone is a shared zone, then the corresponding PPE will get disabled.
• Hardware PPE traps would be seen.
• HW Fault Trap: Count #; The amount of traffic drop can be found in the "count" of the HW fault trap message.
• PFE disable action is expected and if that happens, FPC reset is required to bring it online.

This is typically a transient hardware issue, so no RMA is required.

• If this alarm is raised, at least one PPE zone would be disabled, which in practice does not impact performance (each LU chip contains 16 PPEs, and each PPE uses 24 zones). In this state, the affected FPC continues to operate normally without any performance degradation (with one PPE zone disabled out of 384) and the alarm can be cleared off manually by using the clear chassis fpc errors command.

FPC restart during a maintenance window should clear this error and allow all PPE zones to be used again.

Recommended action

• No action is needed
• If multiple PPEs are disabled, you may schedule a maintenance window to restart the FPC to clear the error
• If PFE disable action was performed, you may schedule a maintenance window to restart the FPC.

Exception

• If you are running Junos software without the fix for PR1652416, you may need to restart the FPC when encountering this error - see also KB32162 [juniper.net]

If the issue is seen repeatedly even after restarting the FPC, contact Support for further investigation and assistance.

2024-03-20 : Article Created

2024-12-03: Minor changes on the Symptoms.