Two drives in my Drobo were reporting errors. I replaced the 1.5Tb drive with a spare 1.5Tb drive and the rebuilding completed.
I replaced a 1Tb drive with a 2Tb drive and rebuilding started. The last time I looked, it had 43 hours to go… then I went out.
When I came back, none of the Drobo volumes were available, and Drobo Dashboard says that no Drobos can be found.
I don’t get any audible warnings when I unplug the drobo from its USB cable, nor when I plug it back in again.
I’ve rebooted my system, but that made no difference.
It seems as if the Drobo has stopped servicing its USB port. I’m wondering if turning it off/on might correct this, but is that a good idea during a rebuild (which I can hear proceeding, with the flashing yellow/green lights) ???
[quote=“Docchris, post:2, topic:39719”]personally i would leave it a couple of days - certainly while you can hear the disks working and they are still warm to the touch
[/quote]
This will be an interesting experiment in withdrawal symptoms… all of my music is on the Drobo, so I can’t play it on either my PC or my (network connected) HiFi.
I’ve got fallback copies of everything that was on the Drobo, but I prefer not to tinker around with everything, only to have to put it back again in 46 hours time (it will probably run faster now I’m no longer accessing the data).
I forgot to mention:
I use a cheap stethoscope that I bought off Ebay - about £3. It’s great for working out where noises are coming from, or if something (which has moving parts) is still active.
For example, at the moment I can hear the disks spinning in my Drobo, but cannot detect any other activity. However, the stethoscope reveals continual very quiet seek sounds - the disks must be seeking over very short spans…
For these mission critical 24/7 100% availability applications, it is best to have a second Drobo or other similar redundant backup ready for instant fail over
The 43 hours was probably about right - figure 24 hours per terabyte of data stored as a rough number.
In the event of a sudden power loss - like you are trying to “reboot” it - in principle the Drobo has an internal battery backup that is supposed to get the array in a consistent state before powering down. And we’ve seen many reports that it really works :). On the other hand, I would not tempt fate either, so +1 to Brandon.
It might be interesting if you elaborated on “Two drives in my Drobo were reporting errors.”…
Sounds to me like a multiple disk failure train wreck in progress? Let’s hope for the best.
[quote=“NeilR, post:7, topic:39719”]It might be interesting if you elaborated on “Two drives in my Drobo were reporting errors.”…
Sounds to me like a multiple disk failure train wreck in progress?
[/quote]
I’ve had a few incidents where Drobo has reported “Cannot protect your data”, and then it went back to normal a couple of minutes later. I sent in the diagnostics, and was told to replace the drives in Slot 0 and 1. Then I had another occurrence, and I noticed the slot 1 light turn red for about a minute.
So, I replaced slot 1 with the spare drive, and purchased a 2Tb drive to replace the 1Tb drive in slot 0.
The rebuild that happened when I replaced the drive in slot 1 completed this morning, so then I replaced slot 0 with my new 2Tb drive. The rebuild from this is proceeding.
However, the Drobo stopped servicing its’ USB port, so now I have no access to the data, and I can’t tell how long the rebuild will run. Another 40 hours, perhaps.
I’m just going to wait it out, but at this point, I don’t think there’s a problem. The lights tell me that a rebuild is in progress, and the capacity lights indicate that the 2Tb drive that I swapped in this morning has increased my capacity (usage has dropped from 75% to around 55-60%). I just need to be patient.
Swifty, so you are on the 2nd drive replacement. Not so bad.
You may have already done this, but if it were me, I would have exercised the drives thoroughly (at least 3 sets of read or write-zero runs across the entire disk) before using them. I do that anyway with all my drives but in a case like yours I think it would be super-critical since you have 2 suspect drives.
Of course, if the first one got through the relay, that should be about the same as a standard burn in so it is all probably moot now.
I worry about all the reports of very early drive death I see on all the drive models. It makes me leery of doing multiple back to back rebuilds with new drives on a single redundancy array.
[quote=“NeilR, post:12, topic:39719”]You may have already done this, but if it were me, I would have exercised the drives thoroughly (at least 3 sets of read or write-zero runs across the entire disk) before using them.
[/quote]
I normally keep a spare disk, and that disk is normally kept in the caddy on my desk, where it is used as temporary space, so it is quite literally a “hot spare”.
But in this case, with two drives going bad, I just ran out of time for this precautionary measure. Now I’ve run out of cash, and will have to wait until next month before I can get another “hot spare” to drop into the caddy.
At the moment, the caddy contains the old drive from slot 0 of the Drobo, which seems still to be OK, but it has exhausted its supply of spare blocks, so probably is not long for this world.
