Mac Pro, ATTO R680, iStoragePro 8-Bay (6gb/sec) = RAID 6 and Confidence Crisis. What Now??

[David Roth weiss]

a) Build new Raid 6 disc group. Reformat volume as big single volume and press on?

Run Disk Warrior right after building. Test, test again, then test some more. If your tests fail then consider moving to plan “e” below.

e) Get rid of enclosure and card and try something else that will give me size, fault tolerance, and support video editing in FCP 7/X, etc.

David Roth Weiss
Director/Editor/Colorist
David Weiss Productions, Inc.
Los Angeles
https://www.drwfilms.com

Don’t miss my new Creative Cow Podcast: Producing Episodic TV with “24” Producer Michael Klick:
https://library.creativecow.net/weiss_roth_david/Podcast-Series-1_Michael-Klick/1

POST-PRODUCTION WITHOUT THE USUAL INSANITY ™

Creative COW contributing editor and a forum host of the Business & Marketing and Apple Final Cut Pro forums.

[Mike Masse]

Thanks for your perspective. When you say “test, test, test” what kind of test are you thinking of? Real world use? DiskWarrior? Apple Disk Utility?

[David Roth weiss]

[Mike Masse] “When you say “test, test, test” what kind of test are you thinking of?”

Just wanted to encourage you to avoid getting a big real-world hurt while you still have uncertainties about your configuration. So long as you have a backup real-world testing is just a loss of time anyway, but I’d really run that setup through the mill now. Try to break it… And, like a lot of us reviewers do, if all seems well once you get it back up and do your initial testing, remove a drive intentionally so you can test the RAID’s abilities in a one and/or two drive failure.

What you don’t know at this point is exactly what created your issue, and whether your hardware really plays nicely together. Buying RAID parts separately does require some diligence on the part of the buyer that is already done for you when buying a more expensive pre-configured RAID.

David Roth Weiss
Director/Editor/Colorist
David Weiss Productions, Inc.
Los Angeles
https://www.drwfilms.com

Don’t miss my new Creative Cow Podcast: Producing Episodic TV with “24” Producer Michael Klick:
https://library.creativecow.net/weiss_roth_david/Podcast-Series-1_Michael-Klick/1

POST-PRODUCTION WITHOUT THE USUAL INSANITY ™

Creative COW contributing editor and a forum host of the Business & Marketing and Apple Final Cut Pro forums.

[Mike Masse]

Status update:

Decided that the 18TB volume created with creating a RAID 6 volume from 8 x 3tb drives was simply too new – it hasn’t been possible to do this (for only a few thousand dollars), and OS/X (10.7.1) hasn’t had a enough time in real world. A 18 TB puts a huge number of eggs in one basket.

4 x 3tb Raid1 volumes has following benefits:
– gives me enough storage for a while (year?)
– keeps volume size completely within the typical range
– creates discs which could be pulled and (as I understand it, but need to test) could be read by the OS directly, without a raid car.
– means some juggling of data, but acceptable.
– has the advantage in FCPX (don’t laugh..) of not mounting some volumes and keeping event catalog size managable without using the steve marting “.dmg” trick.

OK… so moving a head with this, but one of the raid groups (mirrored pairs of discs), shows status onlline but gives me a persistant blinking green dot.

During the build phase this group, there was red dot (instead of the red exclamation point on the other groups during rebuild).

Current Green Dot:

Detail:

In Context:

Current plan is to wait until Monday, and call ATTO Tech Support. The tech support team as been well above average, but the “ATTO ConfigTool Help” does not have explanation for these dots.

Comments on wisdom of RAID 1 welcome.
– fewer eggs in each basket
– drives may be more recoverable without striping and parity
– forces me to keep data in 2.5tb chunks max, which means backup onto 3tb deskstars via drive dock is pretty easy way to get a backup off site.

-m

[Steve Modica]

Atto had a bug with one of the latest driver revisions where the raid would come up with a drive degraded (even tho it wasn’t). We caught this in testing. They also had a bug with NCQ that would take a drive out.

When we see a drive fail in an unusual way, we get them back to Hitachi. We’re an official OEM and they send it through an analysis to tell us what happened. Our two most recent cases were faulty head actuator and a bad midplane in the chassis (power sagging).

Steve Modica
CTO, Small Tree Communications

[Mike Masse]

Steve –

Interesting (if frightening) link to NCQ (native command queuing) – did “SMART” data show any damage?

Your case sounds a little like what happened to me:
– drive showed up as degraded.
– “off to the races” (i.e. would not longer mount / chaos and raid group loss ensued).

Hitachi has been great about replacing / checking drives for our company, but there has always been an onboard S.M.A.R.T. data confirmation of a drive problem – and that has always been a threshold for drive exchange / hitachi tech support (although I’ve never really tested the boundaries before).

Michael

[David Roth weiss]

[Mike Masse] ”
Comments on wisdom of RAID 1 welcome.
– fewer eggs in each basket
– drives may be more recoverable without striping and parity
– forces me to keep data in 2.5tb chunks max, which means backup onto 3tb deskstars via drive dock is pretty easy way to get a backup off site.
“

Bad idea!

RAID 5 and RAID 6 are used almost universally by most video facilities because those two configurations are considered to provide the most efficient mix of throughput, storage space, and protection.

RAID 1 is a huge waste of the most expensive storage space you’ll ever own. Yes, bare hard drives are cheap, but put eight in an enclosure and tie that to an expensive ATTO controller, and it becomes the most expensive storage you’ve got. That storage is better used for primary video storage and increased throughput than wasting 50% for backup in a RAID 1 config. Firewire backup or just chucking three massive SATA drives into the three empty slots in a MacPro is a much more cost efficient way of creating backups.

And, the concept of creating those 2.5Tb manageable chunks or partitions sounds much better than it actually works in reality. That becomes a royal pain in the butt. One big RAID 5 or 6 that works flawlessly is the way to go, and you shouldn’t over think it just because you’re having issues now. Once you have the proper hardware and it’s setup properly you don’t even have to think about it. When you break it into chunks you have to manage those chucks, which is just unnecessary labor.

David Roth Weiss
Director/Editor/Colorist
David Weiss Productions, Inc.
Los Angeles
https://www.drwfilms.com

Don’t miss my new Creative Cow Podcast: Producing Episodic TV with “24” Producer Michael Klick:
https://library.creativecow.net/weiss_roth_david/Podcast-Series-1_Michael-Klick/1

POST-PRODUCTION WITHOUT THE USUAL INSANITY ™

Creative COW contributing editor and a forum host of the Business & Marketing and Apple Final Cut Pro forums.

[Bob Zelin]

Bob Zelin
maxavid@cfl.rr.com

why don’t you contact me Mike. Why don’t you stop playing engineer, and call someone that knows the R680 and CI chassis very well.
We will reconfigure your RAID, and it will all work perfectly when we are done.

What will I charge you for this ? – ZERO.

Just email me. Stop trying to be a f#$%ing hero. I am the hero.
Tomorrow is Monday 9/26. I expect to hear from you.

Bob Zelin

[David Martin]

Nice post Bob 🙂

David Martin
Application Engineer ATTO Technology

[Simon Blackledge]

We nearly went for a same setup but a 16bay but had issues with getting the storage array in UK.

Please keep this updated.

Are large volumes like this in Lion no go currently ?

Atto are always great with support for me.

Please post updates.

ps.. Bobs ya man.. trust 😉

s