Panic: ufs_dirbad: mangled entry

[trh411]

My FreeBSD-9.1-RELEASE system panic'd the other night during my nightly rsync backup job. Here is the pertinent info:

panic: ufs_dirbad: /mnt/backup: bad dir ino 5060196 at offset 0: mangled entry
cpuid = 1
KDB: stack backtrace:
#0 0xffffffff809208a6 at kdb_backtrace+0x66
#1 0xffffffff808ea8be at panic+0x1ce
#2 0xffffffff80b2774f at ufs_dirbad+0x4f
#3 0xffffffff80b28e49 at ufs_lookup_ino+0x6a9
#4 0xffffffff8096ceb8 at vfs_cache_lookup+0xf8
#5 0xffffffff80c68880 at VOP_LOOKUP_APV+0x40
#6 0xffffffff80974554 at lookup+0x464
#7 0xffffffff80975669 at namei+0x4e9
#8 0xffffffff80986993 at kern_statat_vnhook+0xb3
#9 0xffffffff80986b55 at kern_statat+0x15
#10 0xffffffff80986c1a at sys_lstat+0x2a
#11 0xffffffff80bd7ae6 at amd64_syscall+0x546
#12 0xffffffff80bc3447 at Xfast_syscall+0xf7
Uptime: 14h4m38s

The system restarted okay, and after fsck repaired the inconsistencies on the / filesystem, the system completed the reboot normally. Since the panic was caused by a mangled directory on the external USB drive, I ran fsck on it to clean up any problems. This completed normally. I then re-ran my nightly rsync backup job, and this too completed normally. I have had no problems since. I have run fsck on the drive filesystem each day since the panic and the filesystem has checked out clean each time.

I ran various smartctl commands against the external USB drive and although the overall health of the drive was reported as okay, I did see one item in the drive attributes that concerned me:

Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME          FLAGS    VALUE WORST THRESH FAIL RAW_VALUE
  1 Raw_Read_Error_Rate     POSR--   100   253   006    -    0
  3 Spin_Up_Time            PO----   098   098   000    -    0
  4 Start_Stop_Count        -O--CK   092   092   020    -    8889
  5 Reallocated_Sector_Ct   PO--CK   100   100   036    -    0
[color="Red"]  7 Seek_Error_Rate         POSR--   066   060   030    -    4383733[/color]
  9 Power_On_Hours          -O--CK   100   100   000    -    848
 10 Spin_Retry_Count        PO--C-   100   100   034    -    0
 12 Power_Cycle_Count       -O--CK   100   100   020    -    435
187 Reported_Uncorrect      -O--CK   100   100   000    -    0
189 High_Fly_Writes         -O-RCK   100   100   000    -    0
190 Airflow_Temperature_Cel -O---K   067   041   045    Past 33 (Min/Max 33/33)
192 Power-Off_Retract_Count -O--CK   100   100   000    -    315
193 Load_Cycle_Count        -O--CK   092   092   000    -    17455
194 Temperature_Celsius     -O---K   033   059   000    -    33 (0 18 0 0 0)
195 Hardware_ECC_Recovered  -O-RC-   081   057   000    -    11118736
197 Current_Pending_Sector  -O--C-   100   100   000    -    0
198 Offline_Uncorrectable   ----C-   100   100   000    -    0
199 UDMA_CRC_Error_Count    -OSRCK   200   200   000    -    0
200 Multi_Zone_Error_Rate   ------   100   253   000    -    0
202 Data_Address_Mark_Errs  -O--CK   100   253   000    -    0

Based on my research the seek error rate can be caused by servo or head issues, or by high temperature problems. It is also approaching its threshold value. Should I be thinking of replacing the drive based on this data?

 

[giorgiob]

I have a similar problem with a new external USB drive. I bought it a few weeks ago and I use it as an external backup drive attached to a raspberrypi running FreeBSD 11. Unfortunately the filesystem gets corrupted pretty often (once every two, three days) and most of the backup must be recreated. So it is basically unusable. I wonder if it is a hardware or a software problem: right after bringing the drive home, I ran a complete check for bad sectors and no errors were detected.