This article describes a scenario where the FPC encountered a reboot and the last reboot reason is CPU watchdog reset. This article also explains the meaning of the below panic log message.

<3>EDAC MPC85xx MC1: Err Detect Register: 0x80000008
Kernel panic - not syncing: UE mpc85xx_mc_err on mc#1csrow#0channel#0 (csrow:0 channel:0page:0x10000d offset:0xf00 grain:8)

Log messages:

Dec 25 01:01:45.243 lab.qfx ppmd[9147]: %DAEMON-3-PPMD: Connection Shutdown/Closed with PFE: fpc0

Dec 25 01:03:15.000 lab.qfx : %USER-5: fpc0 fpc0 logger: Last Reboot Reason: CPU Watchdog Reset

Panic entries logged in  /var/log/mtd-oops-panic-fpc0-2023-12-25-01_03
 

<3>EDAC MPC85xx MC1: Err Detect Register: 0x80000008

<3>EDAC MPC85xx MC1: Captured Data / ECC:   0xdeadffef_deadbeef / 0xee

<3>EDAC MPC85xx MC1: Err addr: 0x10000df00

<3>EDAC MPC85xx MC1: PFN: 0x0010000d

<4>EDAC MC1: 1 UE mpc85xx_mc_err on mc#1csrow#0channel#0 (csrow:0 channel:0 page:0x10000d offset:0xf00 grain:8)

<0>Kernel panic - not syncing: UE mpc85xx_mc_err on mc#1csrow#0channel#0 (csrow:0 channel:0page:0x10000d offset:0xf00 grain:8)


There are no cores recorded on the device

There are no hardware errors or filesystem errors noticed.

No DIMM / memory error logs on the device

Kernel panic is a system crash or a kernel crash that prevents the system from continuing its normal operation. The crash occurs when the kernel encounters an unrecoverable error or a condition it cannot handle. In the above panic logs, the device encountered ECC error condition which caused the FPC to crash.

Since the FPC crashed, the FPC was unresponsive momentarily. This resulted in watchdog timer not updated in a timely manner by the Routing Engine. When this happens, the watchdog message is placed in the syslog, and the connection between Routing Engine and the FPC is reset. The FPC reboot reason hence shows up as watchdog reboot.

Dec 25 01:01:50 ch_connection_shutdown: Destroying the IPC pipe
Dec 25 01:01:50 fpc_disconnect_generic: fpc 0 state Online cargs 0x9ce6680 clean_shutdown 0, offline_reason=Reconnect


This is a transient issue and if the panic error log occurs again, please raise a case with JTAC for further investigation.

2023-12-26 : Initial Publication
​​​​​​​2024-07-29 - Visibility modified