[conspire] Happy fun HD failure

Rick Moen rick at linuxmafia.com
Sun Jun 6 11:44:32 PDT 2010

Quoting Rick Moen (rick at linuxmafia.com):

> Every bit of data that matters on linuxmafia.com is on a mirror pair of
> hard drives (SCSI HDs /dev/sdb and /dev/sdc, 18 GB each).  sdb just 
> failed.
> Hilarity ensues?

I still haven't replaced the failed half-mirror, but it struck me that
having an absolutely current backup set was more urgent.  Most other
measures I could take would require shutting down the machine, and
backup comes first.  

I'm a big, big fan of offsite backup, but that's slow and I'm in a
hurry.  So, as of yesterday, I'm the owner of a nice shiny 2 TB hard
drive in a 480 Mbit/sec-capable USB 2.0 enclosure.  Man, that's a lot of
space.  Here's the unit I bought, a Hitachi model HSD2000 "SimpleDrive":

Creating the inode tables ran _overnight_, which is a first for me:

# mkfs.ext3 -v /dev/sdd1
mke2fs 1.40.3 (05-Dec-2007)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
244203520 inodes, 488378000 blocks
24418900 blocks (5.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=0
14905 block groups
32768 blocks per group, 32768 fragments per group
16384 inodes per group
Superblock backups stored on blocks:
        32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
        4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968,
        102400000, 214990848

Writing inode tables: done
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done

This filesystem will be automatically checked every 39 mounts or
180 days, whichever comes first.  Use tune2fs -c or -i to override.


I'm not sure the USB port on a 2001-era VA linux Systems model 2230
motherboard is capable of 480 Mbit/sec data transfers, but in any event,
local file copying seems plenty fast.

After I'm done with that (making a safety copy of all current data),
I'll have to see about fixing that degraded mirror pair.

More information about the conspire mailing list