[Jeffrey French]

The short story… the RAID is rebuilt and working.

The long story is beyond my ability to explain, but I’ll see if my friend (who fixed it) can do so in this forum.

Basically, He used some software to clone each of the four (2TB) drives, individually, bit by bit. Two of the drives cloned quickly and without trouble. A third was basically fried. And the fourth had serious issues, but after (literally) weeks of progress creeping along (and noting bad sectors), it managed to complete the clone.

With three fresh, cloned drives, the controller did the rebuild.

I’m scouring my FCP projects now to locate the bad data/corrupted files. There are some (including the VERY last file I was working with before the crash), but not much. The files that are turning up glitchy are easy enough for me to replace with my backups. As mentioned, I have all raw footage and other elements, but no backups of edited projects.

I also have a SECOND set of drives with the same full RAID array backed up, now. In all likelihood, I’ll set up a new array as RAID 6, in addition to backing up regularly.

I learned this lesson the hardest way possible.

[Jeffrey French]

No, no. I’m saying that, in addition to a single brand new drive I have on hand to replace any dead drives, I also have four spares that have already been used in another RAID. I was wondering if it matters whether or not the replacement drive is formatted in any particular way before hot swapping, or if a drive previously used in a RAID might confuse the controller.

In the meantime… I did a hot swap this morning. I pulled the bad drive out and replaced it with a brand new drive (never previously used or formatted). The light on the enclosure is blue, but the GUI says “Rebuilding 0%” and has made no progress for many hours.

[Jeffrey French]

Thanks for your response, Bob. I didn’t explain that very well, sorry.

I don’t know that Drive #4 ever failed. I’ll include some pictures of what I’m seeing. Drive #4 has “FAILED” next to “Reallocated Sector Count” in the SMART window, but it’s still mounted. My guess is it’s in trouble, though.

Drive #1 is dead though, I think. As of this morning, it’s no longer trying to rebuild. The GUI now shows it’s offline. And for the first time today, the red warning light on the exterior case (for the entire RAID, not just the Drive #1 enclosure) was blinking and beeping.

I have a brand new, unformatted disk of the exact same make, size, and speed sitting here. I also have a set of four drives that were previously used as an array in this same box, but I don’t need that data any longer. I’ve never hot swapped a dead drive before, so I don’t know if it requires any formatting in advance, or if there’s a benefit to using one over the other for hot swapping.

And as I’ve never done this before, I’m a little freaked out, and worried that I could somehow do more damage by trying to hot swap a new drive ( in the #1 bay), if the rebuild doesn’t work.

The issue with Drive #4 is that it shows 200+ “bad sectors found and repaired” (Drive #3 also has this problem), as well as the “Reallocated Sector Count” issue in the SMART window. But I don’t know if that means the drive is dead. Yet.

[Jeffrey French]

Yes, the EDL should work. The problem there is that the grading work is being done off site, and I’d hoped to export just the footage he’d need via MM to a small portable drive. The footage and other media for this project totals just over 5TBs and lives on a fairly non-portable drive right now (G SPEED esPRO with its mini-SAS RAID controller card).

I’ve got word out to Singular Software (DualEyes/PluralEyes), and a few other places. Fingers crossed. Thanks.

[Jeffrey French]

Michael: I exported one of the offending clips from the timeline as you said, then re-imported that clip and set it on the timeline. When I right click on that file, Media Manager appears to work fine.

Of course, the problem is that I have 1044 cuts on this timeline (I exported an EDL to see the exact count), and doing this same fix for roughly half of those clips would take more time than I’d like to spend on it.

Another work around we considered was exporting the entire timeline, re-importing it to a new timeline, then getting the cut tool out and chopping that 90 minute file into 1044 new clips. That would seem like a lot of work, too.

Plus, I just don’t like knowing how or why this is happening, and would like to prevent it from happening again.

[Jeffrey French]

Thanks, Michael. Yes, after reading lots of “Error 34” threads on this and other sites, I checked to make sure all the clips had reel names. They do.

And yes, I understand the colorist doesn’t really need the audio, but that doesn’t seem to be related to the problem. Selecting a single, video-only clip – one of the offending DualEyes clips – by itself, right clicking, and clicking “Media Manager” produces the same “Error 34” message.

[Jeffrey French]

Ben, thanks for the informative response. Just to be clear… you’re not using the G-SPEED es enclosures on a MacBook Pro, right? And you’re using the two G-Qs without redundancy (RAID 0), or with? What sort of speeds are you sustaining, and what type of video are you working with?

I recently purchased the CalDigit single port adapter, for use with a single eSATA drive. Looks like I’d need one of those Firmtek dual port Express Card adapters to use those drives together though, right?

I’ve also been looking at these SATAVaults from Coolgear…

https://www.cooldrives.com/usb2-sata-mini-raid-system-sata-raid.html

Not quite as portable, though.

Thanks!

[Jeffrey French]

riccardo – did you ever find an answer to your question? I’m getting the same type of error an error while “invoking the effect plug-in” (in my case, with a different effect, “tint”).

This is a first for me.