here's the scenario: a few days ago i have been noticed, that one of the hard drives inside my lacie 5big has died. which is fine for me, since i have configured it suitable to my personal needs and without any data loss. this article is about my personal story, how i decided for a replacement and to configure the replacement drive using terminal on os x.


and then there was silence

i use a lacie 5big (thunderbolt 2 version) as additional storage and time machine backup for my imac. i configured two hard drives as raid 1 (mirrored) and three drives as raid 5 (striped). and boom, one of the raid 1 drives stopped working. smartreporter (available on the app store) sent me an email notifying me about the crashed hard drive.

statistics please

manufacturer: seagate
model: ST2000DM001-1CH164
size: 2 TB
firmware: cc47
date of manufacturing: 12/2013
start/stop count: 5146
time powered on: 2235 hours
power cycle count: 3019
power-off retract count: 2823
lifetime writes: 2354 GB
lifetime reads: 363 GB

diskutil's output for the affected raid set

[]$ diskutil appleraid list
AppleRAID sets (2 found)
...
===============================================================================
Name:                 Datenlager
Unique ID:            5C756CEC-C39D-11E5-8597-D2031A563991
Type:                 Mirror
Status:               Degraded
Size:                 2.0 TB (2000054943744 Bytes)
Rebuild:              manual
Device Node:          disk10
-------------------------------------------------------------------------------
#  DevNode   UUID                                  Status     Size
-------------------------------------------------------------------------------
-  -none-    B492AEEC-C39A-11E5-8BCF-716C901F17A3  Missing/Damaged
1  disk9s2   7529D822-C39D-11E5-905A-D2031A563991  Online     2000054943744
===============================================================================

please note the status degraded and the missing/damaged disk.

which hard drive is best for me?

before i start with this, a brief disclaimer: don't expect a suggestion from me on what you should buy or not, i will not answer such questions.

the guys at backblaze do a pretty good job in publishing their own statistics each quarter. the most recent statistics by the time of this blog post are available on their blog, as well as an interesting page about (their) hard drives. i'm pretty sure you'll find other statistics and benchmarks using your preferred search engine.

the hard drive i chose is double the size of the current, even if i don't benefit from the additional space yet. but in the long run i will, because by the time i have to replace the other drive from this set, i will choose one with the same (double) size again and voilà, the raid set's size expands.

the replacement process

was done in a few minutes, since the 5big allows an easy access to each hard drive. in terminal on os x identify the new hard drive using diskutil list; in my scenario it's disk8. and add it to the correct raid set (where /Volumes/Datenlager/ is the mounted volume):

[]$ sudo diskutil appleraid add member disk8 /Volumes/Datenlager/
Started RAID operation on disk10 Datenlager
Unmounting disk
Repartitioning disk8 so it can be in a RAID set
Unmounting disk
Creating the partition map
The added target disk is larger than the RAID set. This will result in unused disk space on the target disk "disk8s2"
Adding disk8s2 to the RAID Set
Finished RAID operation on disk10 LaCie

next i removed the faulty hard drive:

[]$ diskutil appleraid remove B492AEEC-C39A-11E5-8BCF-716C901F17A3 /Volumes/Datenlager/
Started RAID operation on disk10 Datenlager
Removing disk from RAID
Finished RAID operation on disk10 Datenlager

rebuilding

while the new hard drive will be fully implemented, it's status is called rebuilding.

#  DevNode   UUID                                  Status     Size
-------------------------------------------------------------------------------
0  disk9s2   7529D822-C39D-11E5-905A-D2031A563991  Online     2000054943744
1  disk8s2   B83C6F54-C39C-11E5-B612-AE24DB915C80  2% (Rebuilding)2000054943744
===============================================================================

once the rebuilding has finished, the status of the raid set is online again.

===============================================================================
Name:                 Datenlager
Unique ID:            5C756CEC-C39D-11E5-8597-D2031A563991
Type:                 Mirror
Status:               Online
Size:                 2.0 TB (2000054943744 Bytes)
Rebuild:              automatic
Device Node:          disk10
-------------------------------------------------------------------------------
#  DevNode   UUID                                  Status     Size
-------------------------------------------------------------------------------
0  disk9s2   7529D822-C39D-11E5-905A-D2031A563991  Online     2000054943744
1  disk8s2   B83C6F54-C39C-11E5-B612-AE24DB915C80  Online     2000054943744
===============================================================================

what i find very nice is: during the addition of the new member, the key rebuild changed it's value from manual to automatic.

conclusion

all in all it was a very straight and easy process. no voodoo magic, or ritual dance around the bonfire at full moon necessary. thanks to apple and it's powerful command line tools like diskutil. this experience made me even more convinced that i did nothing wrong with my decision for this setup.


note: all uuid's in this article have been replaced with other values due to security reason.

main picture by pierre rougier, license