Fail Recovery

Of course, one of the reasons you purchase a RAID system is for fault tolerance.  For my fail recovery test, the Media Vault was configured for RAID 1.  I had previously copied about 60GB of data and was copying an additional 4GB when I pulled the top drive (2) out.  Within about 8 seconds or so, the health indicator light turned red. Shortly thereafter, the light bar adjacent to Drive 2 went out.  

As expected, the Media Vault remained online, and my file copying continued without a hitch.  I logged into the admin console and the only notice on the home page was “alerts – view log”.  But in viewing the log, I found confusing information.  The log indicated that both drives failed (Figure 14).

Figure 14: Alert log showing both drives failed

However, the “Disks” page reflected what was really going on (Figure 15).

Figure 15: Disks page showing disk 2 unavailable

Since the Media Vault doesn’t “officially” support hot swap, I shut it down, re-inserted Drive 2, re-booted and found the Drive 2 “light bar” showed purple.  According to the User's Guide this means that "a hard drive has been installed, but has not been initialized". When I logged into the device, the RAID still showed as degraded on the disk page.

I clicked on the small “fix” icon and the mirror started rebuilding with an estimated 10 hour of rebuild time (which actually turned out to be 6-7 hours). Interestingly, shortly after the rebuild started, the health indicator return to normal blue status, as did the light bar next to Drive 2! This would indicate that the drive was back to fault tolerant mode. However, a check of the disk status page showed that the array was only 5% rebuilt!  This is a very misleading situation and HP should fix this as soon as possible.