Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Each his own, but there's plenty of consolidation and simplification to be done if you want which might reduce risk of data loss without impacting performance. For example, running your massive iTunes library on a RAID0 array is unnecessarily risky and none of that content can benefit from the added performance RAID0 provides over a single disk.

I'm not sure how big your active working data set is, but if you haven't already, I would put as much as you can onto SSD when working on a project and then performance of the rest is less important and you can get that off RAID0 and onto something more reliable as well.

Anyway, the reason were having this discussion is because you don't think you could adapt your storage solution to fit with the constraints of the new Mac Pro... And I disagree. I have nearly he exact same data storage types and requirements but due to the way I've set it up, the new Mac Pro will not require me to change anything. It's all up to you.


Starting with the iTunes library - drive dies, replace the drive with the spare I bought at the same time I bought the other drives, restore from external raid 5 - mission accomplished.

Scratch drive - drive dies, replace drive with the spare, restore from external raid 5 - see above.

The working set won't fit onto the SSDs - that was the original plan, (1 for OS & apps, 1 for working set) until I really started to dig around into my runtimes - I'm looking at about 3/4 of a Tb, so it isn't really an option.

I can adapt my storage solution to the nMP - it's just that it will cost about $1,800 to do it, by all means, if you can show me a cheaper way, I'm listening. Keep in mind the following:

Boot drive - 311Gb of OS & applications
Scratch drive - 2.5Tb of data
iTunes drive - 3.48Tb

Backup system for all 3 logical drives. And don't forget all of the peripherals while your are at it.
 
Starting with the iTunes library - drive dies, replace the drive with the spare I bought at the same time I bought the other drives, restore from external raid 5 - mission accomplished.

Scratch drive - drive dies, replace drive with the spare, restore from external raid 5 - see above.

The working set won't fit onto the SSDs - that was the original plan, (1 for OS & apps, 1 for working set) until I really started to dig around into my runtimes - I'm looking at about 3/4 of a Tb, so it isn't really an option.

I can adapt my storage solution to the nMP - it's just that it will cost about $1,800 to do it, by all means, if you can show me a cheaper way, I'm listening. Keep in mind the following:

Boot drive - 311Gb of OS & applications
Scratch drive - 2.5Tb of data
iTunes drive - 3.48Tb

Backup system for all 3 logical drives. And don't forget all of the peripherals while your are at it.

Well, I would start with 1TB of SSD in the new Mac Pro (assuming it's offered) to accommodate your workflow (that will improve your performance more than any of this other stuff). Otherwise I would go to a Promise J4 TB enclosure for $375 to have your SSD storage external. Then I would put the scratch on a single fast 4TB drive (USB3) (likely faster than your RAID0). Then I would ditch the iTunes drives and just maintain your library on the RAID5 array which is already protected from drive failure.

So for $375 for the J4 (if buying SSD in the new Mac Pro is not possible) and another $199 for a 4TB USB3 drive, you're set.
 
I guess my point was that not all your drives need to be inside your Mac for them to meet your requirements. In your situation, whether you move to the new Mac Pro or not, you might benefit from getting your storage house a bit more organized :)... fewer larger drives and/or a NAS might be a good idea... basically all your backups, media files, and archived work could live on a NAS RAID and wouldn't suffer at all. Your large application data could sit on a simple large/fast 3 or 4TB drive that could be made external at the drop of a hat and you could look at moving a minimal working set to SSD to improve your performance.

Not every hard drive you ever purchased needs to be kept until it stops spinning from old age. :p
the flaw with this logic is that if you put everything on one big fast drive, you are putting all of your eggs in one basket.
I've had drives from every manufacturer fail. Some after a couple months, some after 5 years. But no drive is 100% safe to keep your data on.
I always set up all of my systems to have an OS volume and a data volume on separate media.
Besides, your solution ends up with 3 different volumes anyway. Two of which are external. Not sure how that is better.
 
I'll make it less complicated for you......

Raid 5 - external - 1 drive goes down - no data loss, I pull the drive, replace it with the 5th drive I bought when I purchased the other 4, raid rebuilds. The only way I can lose the array is if multiple drives die at the same time. Any 1 drive system will lose everything if it goes down.

Save money by spending $1,200 on 4tb drives to replace working 2tb drives - what is that, Republican math?

No, my iTunes library isn't way too big. It may be too big for your needs, but it isn't for mine - I like having all 4,000 CDs available - it makes the shuffle feature real interesting.

The 2 hard drives take up less space than the wall of cds and dvds that it replaced.

Let me know how that goes when one of your RAID5 drives fail. It's a little overkill for iTunes. I don't know why you think it costs $1200 to buy a 4TB drive. What is that, troll math? You could spend every waking minute listening to your iTunes library, and you probably wouldn't even make it half way through. What's the point? I have around 1000 albums, and it's borderline overkill. Oh wait, this is relevant to the new Mac Pro how?
 
Let me know how that goes when one of your RAID5 drives fail. It's a little overkill for iTunes. I don't know why you think it costs $1200 to buy a 4TB drive. What is that, troll math? You could spend every waking minute listening to your iTunes library, and you probably wouldn't even make it half way through. What's the point? I have around 1000 albums, and it's borderline overkill. Oh wait, this is relevant to the new Mac Pro how?

Probably like the only time I had a raid5 fail - pulled the dead drive, replaced it with another, raid rebuilt itself. No muss, no fuss. But then, I have a lot of experience with Raids.

Which do you think is faster - restoring 3.48 Tb from a backup or re-encoding 3,748 CDs into iTunes. I am betting on a restore myself.

Bare 4Tb drives run just under 300 - If you paid attention to what I actually wrote you would have realized that I currently have 8 2Tb drives (there's that math thing again) to hold my data - if I go to 4Tb drives, I need at least 4 of them, hence $1,200 for 4x4Tb drives. And that doesn't leave room for further expansion.

As far as how long it would take to listen to everything on my iTunes library - 168.4 days. I've been collecting CDs since 1984. I suspect that your collection is nothing more than a collection of "Now That's What I Call Music" compilations. Nothing wrong with that, but what can I say - l like many styles of music, and I have a lot of it.
 
Well, I would start with 1TB of SSD in the new Mac Pro (assuming it's offered) to accommodate your workflow (that will improve your performance more than any of this other stuff). Otherwise I would go to a Promise J4 TB enclosure for $375 to have your SSD storage external. Then I would put the scratch on a single fast 4TB drive (USB3) (likely faster than your RAID0). Then I would ditch the iTunes drives and just maintain your library on the RAID5 array which is already protected from drive failure.

So for $375 for the J4 (if buying SSD in the new Mac Pro is not possible) and another $199 for a 4TB USB3 drive, you're set.

1TB SSD won't accommodate my workflow - I tried that already with a 1Tb hard drive - I didn't build those 4tb Raid0s out of boredom - I ran out of space.

The 4TB USB3 scratch disc isn't very likely to be faster than my current setup - I am averaging 201.7MB/s (1Gb file transfer) on the scratch disc - and that is while the drives are being accessed by other applications in addition to the benchmarking program. The real world numbers I have seen for USB3 are about 40% less than that. Also keep in mind that those numbers drop as the disc is filled - and I am approaching 70% on the scratch disc, and 90% on the iTunes disc.

The raid won't hold both my iTunes library & my backups (4x2TB = 6TB of usable space) and I'd have to buy a ESATA II to USB3 adapter. This moves my USB port requirement to 8. Which means I need 2 Belkin TB docks instead of 1, if I moved to the nMP.

If I move the iTunes to the external raid, it will be at 55% fill, right now. Read/Write speeds will steadily drop as I fill it. And much to portishead's annoyance, I am not done buying music & video - I have 6 discs on the way to the house now (Volunteers/Bark/Long John Silver by Jefferson Airplane, Sunfighter/Baron Von Tollbooth & Chrome Nun by Kanter & Slick, and Call to Arms by Saxon).

I'll still need a back up strategy for the scratch disc. The cheapest would be to move the internal drives to yet another Qx2, which will put me back another $300 (enclosure + esata to USB3 adaptor). Now I am using 9 USB ports if I move to a nMP.

Oh, and 1 more thing.......

All of the nMP connectors are in the back of the can. How many months do you think before TB cables start breaking if someone is spinning it around to plug/unplug items on a regular basis.

As a minimum, I'd want at least 2 spare TB cables and a USB3 cable for when they inevitably break.

But that is just me looking ahead.
 
OK, but in the light of recent successes like the iPhone, MacBook Pro, etc. you should not forget that sometimes Apple stuck its neck out—only to get a stiff neck because there’s some harsh wind blowing. Think of the
  • G4 Cube
  • Newton
  • Pippin
  • Puck Mouse
  • Xserve/Xserve Raid
All have been products where Apple tried something different and the market just said: Meh! Don’t like it because of [put your favorite reason here].


What actually puzzles me is the discrepancy between the direction Apple takes with it’s software and the hardware. Apple prices their Pro application extremely competitive. FCPX, LogicX, Aperture … they are all very cheap and thus target the prosumer market as well as the pro user market. Clearly they want to put their software into as many hands as possible.

The prices of most Macs has been quite stable over the recent years or even went down. And even if the prices did not went down, Apple built successors of much higher quality. Just compare the iBook/Macbook line with the 13" MacBook Pro or Air. Again, I think their intention is to put their hardware into as many hands as possible.

But then comes the new Mac Pro and all the components individual prices point to a higher system price level than before - or perhaps just the same 2500 for an base configuration which needs some external augmentations via USB/Firewire thus making the whole system most like more expensive.

Well, we don’t know the final prices for each new Mac Pro … so it’s all just speculation.

So here is my stupid idea:
After Apple introduces the new Mac Pro as a high end system, they will take the new Mac Pro form factor and offer the so called “new Mac” (without the Pro moniker). It will have a consumer chipset which supports Haswell consumer CPUs, consumer GPUs (one or two), have less Thunderbolt ports but otherwise uses the same gestalt.
This is what I would like to see as the new Mac (or MiniPro or Xmac). Target the system at the 1500-2000 $ price range and it would perhaps be another top seller.

They'll never call a Mac "Mac".
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.