Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Boot off RAID 0 in Mac Pro

HyperX13

HyperX13

macrumors 6502
Original poster
Sep 3, 2009
351
7
Is it possible to created a Raid 0 in Mac Pro and then boot off of it. I can do it with my HP, I cant seem to see how to do in on a Mac Pro. (I have the 2008 Mac Pro).
 
nanofrog said:
Yes.

Look in Disk Utility IIRC. :)
Click to expand...

OK... I must be doing it wrong, because it doesnt boot, but I can access it if I use external disk to boot... gonna look into it some more.. I swore I had this working on my 2006 Mac Pro and Leopard...
 
You create your RAID0 array and use the SL install disk to clone your existing boot partition onto the RAID0 array.

Then you boot with the option key pressed and you should be offered with the option to boot into the new boot partition.

If not you can use disk utility to repair the partition or the permissions.
 
4x sas 15.000 rpm with raid card, as a boot drive ....

Hi all, i'm about to set up my system aswell only i will be using 4 sas drives in raid-0, thats the fastest i can get my boot drive, or am i wrong?

Also with this setup ,will it be a problem that when working with FCP the scratch disk will bassicly be on the same stripped raid set, or will this be no problem because of the high I/O of that stripped array ?

I have an external eSata drive that can also be used as a scratch disk that should be able to go around 150MB/s, but i wonder if this will be necasery since i expect a far higher number from my raid-0.

But i do this mostly to get my set up as fast as i can, also when not doing any editing on my system, i hope it will fly ;-).

I use a 8core mac pro thats 1,5 years old now with the raid card installed.

And before you ask, yes of course i'm backing the whole thing up to on a hourly basis ;-).
 
Data said:
Hi all, i'm about to set up my system aswell only i will be using 4 sas drives in raid-0, thats the fastest i can get my boot drive, or am i wrong?

Also with this setup ,will it be a problem that when working with FCP the scratch disk will bassicly be on the same stripped raid set, or will this be no problem because of the high I/O of that stripped array ?

I have an external eSata drive that can also be used as a scratch disk that should be able to go around 150MB/s, but i wonder if this will be necasery since i expect a far higher number from my raid-0.

But i do this mostly to get my set up as fast as i can, also when not doing any editing on my system, i hope it will fly ;-).

I use a 8core mac pro thats 1,5 years old now with the raid card installed.

And before you ask, yes of course i'm backing the whole thing up to on a hourly basis ;-).
Click to expand...
As far as what the drives can do, no, it's not wrong. ;) So long as you're aware of the issues involved, namely time consumed when something goes wrong (eventually will, not if).

Scratch will benefit from the array, though having adequate memory for the usage will make more of a difference, as it will prevent it from even needing to do so (scratch space was designed to compensate for the lack of memory when RAM was still really expensive). It's an outdated concept now, as RAM is far cheaper.

You can skip the eSATA idea, unless something else that has access to the array is an absolute I/O hog. Then you'd be better off adding additional drives to the RAID card if at all possible.

BTW, is it Apple's RAID Pro card installed? Or a 3rd party model?
 
Hi, ty for the reply , It is the macpro raid card from Apple and i also have 8 gig ram installed, and with the 4 sas drives installed thats all THE Card can handle thats why i wanted the external raid drive with esata for fast external storage to.
 
Data said:
Hi, ty for the reply , It is the macpro raid card from Apple and i also have 8 gig ram installed, and with the 4 sas drives installed thats all THE Card can handle thats why i wanted the external raid drive with esata for fast external storage to.
Click to expand...
I understand with the RAID card. It's custom made for Apple, and unfortunately, it's crap. Slow, expensive, only supports one OS, and has battery problems. In simple terms, it's junk.

That said, if that's what you've got, then use it. ;) :p I'd recommend getting a different card, and selling that one off if possible.

Using an external backup system is the best way to do it. :) Maybe adding offsite storage as well would be better, as that helps you in case of a fire, water damage,... that can take out both the system and the backups. Ugly, but possible.
 
If heart alot about the card being junk too, but tthat was of course after i bought it so, like you said , thats what i have and that's what i'll use ;-).

It's in my machine from day 1 one, `nd i had the battery replaced within 2 months and after that i have had no problems anymore with the card.

How slow it is when i configure the whole thing as a raid 0 i'll have to wait and see, but i only use mac osx so the lack of support for other systems is no problem for me, lucky me huh ;-).

I've got a couple of days on a small thing i'm working on , when thats finished i'm gonna do the set up in raid0 and then i'll see what happens.

Hope it's not to slow ;-).
 
gugucom said:
You create your RAID0 array and use the SL install disk to clone your existing boot partition onto the RAID0 array.

Then you boot with the option key pressed and you should be offered with the option to boot into the new boot partition.

If not you can use disk utility to repair the partition or the permissions.
Click to expand...

what an odd way to do it, why havent they made it easier then that? so basically, you cannot install a brand new OS to a RAID array?

what happens if you buy a MP with two HDDs and want to install the OS to it? that means you will have to erase the disks then format, RAID etcetc. will you need an external drive to clone the single boot drive?

Data said:
Hope it's not to slow ;-).
Click to expand...

hit us up with some benchmarks :D
 
What kind of bechmark program do you want me to use ? ;-).

As for the set up, i was thinking, i'm gonna erase all the disk and build the raid-0 with the Raid configuration program on the SL disk,and then install the os also from the SL install disk, and then put back my old sytem from my time machine backup. This seems like the easiest way to do it , or is there a problem if i try to work like that ?
 
Data said:
What kind of bechmark program do you want me to use ? ;-).
Click to expand...
im not familiar with RAID testing programs.. anything that can test the various bit sizes, e.g. small writes/reads to large writes/reads.

As for the set up, i was thinking, i'm gonna erase all the disk and build the raid-0 with the Raid configuration program on the SL disk,and then install the os also from the SL install disk, and then put back my old sytem from my time machine backup. This seems like the easiest way to do it , or is there a problem if i try to work like that ?
Click to expand...

that would most likely be the easiest way. if thats possible then you answered my last question, thanks :D
 
i hope i'll get to it by the weekend , i will let you know if it worked, i'll give the aja test program a go then for some numbers ;-).
 
Data said:
What kind of bechmark program do you want me to use ? ;-).

As for the set up, i was thinking, i'm gonna erase all the disk and build the raid-0 with the Raid configuration program on the SL disk,and then install the os also from the SL install disk, and then put back my old sytem from my time machine backup. This seems like the easiest way to do it , or is there a problem if i try to work like that ?
Click to expand...
You could try AJA System Test.
 
@dofot9 , i'm still paying for mine too, by the time it's paid in full it will be obsolete again,lol.
 
Data said:
@dofot9 , i'm still paying for mine too, by the time it's paid in full it will be obsolete again,lol.
Click to expand...

ugh so true! but so worth it! i hate taking out loans, mainly because my income is, well, non-existent. cant wait till i have one of these puppies but. let us know how it goes!
 
Xbench result

Hi all, i just did a Xbench here are the results:

Results 214.44
System Info
Xbench Version 1.3
System Version 10.6.1 (10B504)
Physical RAM 8192 MB
Model MacPro3,1
Drive Type APPLE RAID Card
CPU Test 201.32
GCD Loop 327.46 17.26 Mops/sec
Floating Point Basic 164.55 3.91 Gflop/sec
vecLib FFT 132.28 4.36 Gflop/sec
Floating Point Library 314.66 54.79 Mops/sec
Thread Test 823.39
Computation 1124.99 22.79 Mops/sec, 4 threads
Lock Contention 649.32 27.93 Mlocks/sec, 4 threads
Memory Test 200.84
System 290.26
Allocate 380.43 1.40 Malloc/sec
Fill 231.96 11278.46 MB/sec
Copy 294.47 6082.15 MB/sec
Stream 153.54
Copy 146.25 3020.81 MB/sec
Scale 160.00 3305.45 MB/sec
Add 155.01 3301.95 MB/sec
Triad 153.56 3284.98 MB/sec
Quartz Graphics Test 230.82
Line 197.13 13.12 Klines/sec [50% alpha]
Rectangle 267.71 79.93 Krects/sec [50% alpha]
Circle 230.02 18.75 Kcircles/sec [50% alpha]
Bezier 230.01 5.80 Kbeziers/sec [50% alpha]
Text 240.44 15.04 Kchars/sec
OpenGL Graphics Test 94.69
Spinning Squares 94.69 120.12 frames/sec
User Interface Test 340.18
Elements 340.18 1.56 Krefresh/sec


Disk Test 273.92
Sequential 280.19
Uncached Write 409.18 251.23 MB/sec [4K blocks]
Uncached Write 362.19 204.93 MB/sec [256K blocks]
Uncached Read 139.94 40.96 MB/sec [4K blocks]
Uncached Read 519.35 261.02 MB/sec [256K blocks]
Random 267.93
Uncached Write 778.07 82.37 MB/sec [4K blocks]
Uncached Write 324.67 103.94 MB/sec [256K blocks]
Uncached Read 166.94 1.18 MB/sec [4K blocks]
Uncached Read 218.65 40.57 MB/sec [256K blocks]



I don't realy have anything to compare it with so maybe, if you post some to compare this with it might be more usefull ;-).
 
Data said:
I don't realy have anything to compare it with so maybe, if you post some to compare this with it might be more usefull ;-).
Click to expand...

Here's mine:
Code: 
Results	647.34	
	System Info		
		Xbench Version		1.3
		System Version		10.6.1 (10B504)
		Physical RAM		12288 MB
		Model		MacPro4,1
		Drive Type		Areca Mac Pro RAID
	Disk Test	647.34	
		Sequential	565.97	
			Uncached Write	827.42	508.03 MB/sec [4K blocks]
			Uncached Write	616.17	348.63 MB/sec [256K blocks]
			Uncached Read	266.72	78.06 MB/sec [4K blocks]
			Uncached Read	2054.64	1032.65 MB/sec [256K blocks]
		Random	756.05	
			Uncached Write	1136.88	120.35 MB/sec [4K blocks]
			Uncached Write	243.15	77.84 MB/sec [256K blocks]
			Uncached Read	9737.08	69.00 MB/sec [4K blocks]
			Uncached Read	5109.94	948.19 MB/sec [256K blocks]

For reference it's running an Areca ARC-1212 card with write back enabled connected to 4x 300GB WD VelociRaptors in RAID 5.
 
alphaod: those numbers are really nice, but they seem too high, for a set of 4 VR's.

What's the file size used in the tests?

I ask, as it appears that the data is being skewed by the cache (which is nice, so the numbers are accurate for small files). But if you want to know what it does on a file larger than the cache can contain, you can do a couple of things.

1. Increase the file size, which will give you an amalgamation of large file with cache = active.

2. Disable the cache, but the test then only shows what the drives can do on their own.

Your choice. If you can't do the first method with Disk Util, then try AJA System Test (free download).

Hope this helps. :)
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back