Description

The RE (Routing Engine) rebooted unexpectedly with the following message:

reboot after panic: ffs_clusteralloc: map mismatch

Symptoms

This example is an MX960 router.
The following was seen in the syslog with 'show system boot-messages' in the rebooted RE: 

Feb 24 12:59:47  MX960 /kernel: virtual chassis member-id = 255
Feb 24 12:59:47  MX960 /kernel: virtual chassis mode      = 0
Feb 24 12:59:47  MX960 /kernel: Initializing product: 21 ..
Feb 24 12:59:47  MX960 /kernel: Setting up M/T interface operations and attributes
Feb 24 12:59:47  MX960 /kernel: fxp0: bus=5, device=0, func=0, Ethernet address 00:a0:a5:7e:59:42
Feb 24 12:59:47  MX960 /kernel: em1: bus=4, device=0, func=0, Ethernet address 02:00:01:00:00:04
Feb 24 12:59:47  MX960 /kernel: em0: bus=3, device=0, func=0, Ethernet address 02:00:00:00:00:04
Feb 24 12:59:47  MX960 /kernel: acb: acb_late_attach: Initializing silabs clocks to drive 8727 external phy
Feb 24 12:59:47  MX960 /kernel: Timecounter "TSC" frequency 1729008504 Hz quality 800
Feb 24 12:59:47  MX960 /kernel:  Loading Redundant LT driver
Feb 24 12:59:47  MX960 /kernel: ###PCB Group initialized for udppcbgroup
Feb 24 12:59:47  MX960 /kernel: ###PCB Group initialized for tcppcbgroup
Feb 24 12:59:47  MX960 /kernel: ad1: 30533MB <UGB94BPH32H0S1 KCI 2030> at ata0-master UDMA33
Feb 24 12:59:47  MX960 /kernel: ad0: 3831MB <UGB30SFA4000T2 2010108C> at ata0-slave UDMA33
Feb 24 12:59:47  MX960 /kernel: Kernel thread "wkupdaemon" (pid 41) exited prematurely.
Feb 24 12:59:47  MX960 /kernel: Trying to mount root from ufs:/dev/ad0s1a
Feb 24 12:59:47  MX960 /kernel: WARNING: /mnt was not properly dismounted
Feb 24 12:59:47  MX960 savecore: reboot after panic: ffs_clusteralloc: map mismatch
Feb 24 12:59:47  MX960 savecore: writing core to vmcore.0

Solution

This error is due to inconsistency between the cluster map (contiguous blocks) and the block map. It is a form of FS corruption. A cluster is the most basic unit in FS and defines the boot sector. Under FreeBSD 4.9, the same issue has been reported. Refer to FreeBSD Bugzilla – Bug 64406 .

Recommendations:

  1. In general, a faulty hardware occurred in the routing engine such as storage. Create a RMA to replace and monitor for 24-48 hours after installation. Repair analysis may be available, depending on whether the part is EOL.

  2. If RMA unit is unable to resolve the issue, contact JTAC for further assistance.

Modification History

2021-04-30: Minor edit.