[moonmonkey]

I know this is mentioned in other posts (and also dismissed in some posts), but I though I would let you know that this has has a massive effect on my mini.

I bought an external Firewire Disk enclosure and put a Maxtor DiamondMax 10 200 Gig (8 meg cache, 7200 RMB) drive in it.

I installed a new copy of Tiger and used the included the migration tool to copy over all the files from the internal disk.

The entire proccess cost less than 120 USD and the computer absolutly flies, its like a big weight has been lifted off it. This gave me a massive, massive speed boost in nearly all tasks.

If you have a Mac mini (Gen 1) I can really recomend doing this.

[mikemako]

So a 7200RPM disk through Firewire is faster than an internal 4200RPM disk through Ultra ATA?

[tooki]

Yep.

tooki

[asb819]

Good idea! I actually already have a 200GB internal in my old Windows machine. Which external enclosure would you recommend? I found one with good reviews for about 50 bucks on Newegg. Is it possible to just create a mirror image of the original disk on the external, or would I have to actually install the OS and then copy the files? What I'm wondering if it's possible to do something like this: http://www.bixnet.com/ezdattrankit.html

[moonmonkey]

Originally Posted by asb819

Good idea! I actually already have a 200GB internal in my old Windows machine. Which external enclosure would you recommend? I found one with good reviews for about 50 bucks on Newegg. Is it possible to just create a mirror image of the original disk on the external, or would I have to actually install the OS and then copy the files? What I'm wondering if it's possible to do something like this: http://www.bixnet.com/ezdattrankit.html

Don't bother with that, mirroring is built into the Tiger (and Panther I think) installer.

Install the new disk in the drive encloure and connect to your mini
Boot from the OS DVD
Format the new disk as HSF+
Install the OS onto the new drive
After install is completed choose the option to migrate all files from the internal hard disk

The OS will now boot using the external Firewire drive

Now you will have two identical bootable hard disks, you can format the internal one with disk Utility

[moonmonkey]

Originally Posted by mikemako

So a 7200RPM disk through Firewire is faster than an internal 4200RPM disk through Ultra ATA?

Will post benchmarks if anyone is interested.

[ioaz]

I would love it if you could post benchmarks. I am seriously contemplating doing this. I would also like to know what enclosure you used too.

[asb819]

Same here. Would love benchmarks and to know which enclosure as well.

[moonmonkey]

Can someone post the Xbench scores (disk disk access) from an original 1.42 GHz (1 Gig RAM) mac mini so we can compare?

FYI
The enclosure I use is http://www.macpower.com.tw/products/hdd3/m9/m9dx

In the US they are called the NewerTech MiniStack.
http://www.newertech.com/ministack

Any FW400 enclosure would work fine (I just wanted more Firewire ports), although you may need to check if you need a serial or parallel ATA drive for your enclosure (I have serial) parallel may be even faster.

[Simon]

Originally Posted by moonmonkey

Install the new disk in the drive encloure and connect to your mini
Boot from the OS DVD
Format the new disk as HSF+
Install the OS onto the new drive
After install is completed choose the option to migrate all files from the internal hard disk

Much too complicated.

Don't reboot. Just open /Applications/Utilities/DiskUtility, select your mini's internal HD and go to the 'Restore' tab. As a source select the internal HD, as the destination select your new external FW disk. Have the utility format the disk and restore in one step. Go to the system prefs and select the external disk. Done. All preserved. It's as simple as that.

[JazzCatDRP]

I'm debating about doing this as well. I have a 7200 external firewire drive with a lot of music files and whatnot on it. If I do this, will all my files remain intact? It's already formatted as HFS+

[JazzCatDRP]

By the way, my hard drive is this one: http://www.epinions.com/pr-Acomdata_...ay_~full_specs

Could I expect a big enough improvement over my 1.25 5400 40 GB hard drive to do the migration? I noticed it is only a 2 MB buffer.

[moonmonkey]

Originally Posted by Simon

Much too complicated.

Don't reboot. Just open /Applications/Utilities/DiskUtility, select your mini's internal HD and go to the 'Restore' tab. As a source select the internal HD, as the destination select your new external FW disk. Have the utility format the disk and restore in one step. Go to the system prefs and select the external disk. Done. All preserved. It's as simple as that.

I just take every opportunity I can get to reinstall the OS

[moonmonkey]

Originally Posted by JazzCatDRP

By the way, my hard drive is this one: http://www.epinions.com/pr-Acomdata_...ay_~full_specs

Could I expect a big enough improvement over my 1.25 5400 40 GB hard drive to do the migration? I noticed it is only a 2 MB buffer.

Run Xbench Disk test on your mac mini, then run it on the external, then post them up here.

[JazzCatDRP]

Thanks for the help. Here's the results for the mini hard drive:

Results 25.06
System Info
Xbench Version 1.2
System Version 10.4.2 (8C46)
Physical RAM 512 MB
Model PowerMac10,1
Processor PowerPC G4 @ 1.25 GHz
L1 Cache 32K (instruction), 32K (data)
L2 Cache 512K @ 1.25 GHz
Bus Frequency 167 MHz
Video Card ATY,RV280
Drive Type ST940110A
Disk Test 25.06
Sequential 26.91
Uncached Write 56.35 34.60 MB/sec [4K blocks]
Uncached Write 52.50 29.71 MB/sec [256K blocks]
Uncached Read 13.73 4.02 MB/sec [4K blocks]
Uncached Read 25.61 12.87 MB/sec [256K blocks]
Random 23.45
Uncached Write 8.31 0.88 MB/sec [4K blocks]
Uncached Write 45.67 14.62 MB/sec [256K blocks]
Uncached Read 62.23 0.44 MB/sec [4K blocks]
Uncached Read 81.07 15.04 MB/sec [256K blocks]

Here's the results for the external:

Results 37.89
System Info
Xbench Version 1.2
System Version 10.4.2 (8C46)
Physical RAM 512 MB
Model PowerMac10,1
Processor PowerPC G4 @ 1.25 GHz
L1 Cache 32K (instruction), 32K (data)
L2 Cache 512K @ 1.25 GHz
Bus Frequency 167 MHz
Video Card ATY,RV280
Drive Type DMI 00GUA0
Disk Test 37.89
Sequential 37.43
Uncached Write 19.53 11.99 MB/sec [4K blocks]
Uncached Write 57.03 32.27 MB/sec [256K blocks]
Uncached Read 43.13 12.62 MB/sec [4K blocks]
Uncached Read 66.91 33.63 MB/sec [256K blocks]
Random 38.36
Uncached Write 14.41 1.53 MB/sec [4K blocks]
Uncached Write 74.24 23.77 MB/sec [256K blocks]
Uncached Read 85.18 0.60 MB/sec [4K blocks]
Uncached Read 103.66 19.23 MB/sec [256K blocks]

Seems there is a difference...is it enough to make a noticeable difference in speed? Thanks for the help!

[moonmonkey]

Disk Scores in Summary

25.06 Original MacMini (1st release) - internal drive
37.89 Acomdata (HD160U2FE-72) 160 GB - external Firewire Drive
60.06 Maxtor DiamondMax 10 200 Gig (8 meg cache, 7200 RMB) - external Firewire Drive

So the Acomdata will give you an approx 50% speed boost over the standard drive, I would go for it if I were you.

[No Time 4 Love Dr. Jones]

Doesn't this defeat the purpose the Mac mini, though?

[Zim]

Originally Posted by No Time 4 Love Dr. Jones

Doesn't this defeat the purpose the Mac mini, though?

Not if your purpose is to buy a headless iMac (ie. you don't _need_ a Powermac, but don't want the ^#$^ built-in display that comes with anything else... Apple has no other Mac that comes without a screen)

Mike

[JazzCatDRP]

Yep, the reason I have a mini is for the low cost...not space saving features.

[ioaz]

Ok, I'm a little worried here, I just tested my system:
1.42
1GB RAM
80GB hdd - came with mini
and I just scored a 38.34 on Xbench

results

I'm still thinking i'm going to go the external drive route and if I do I will run a benchmark right before and right after to check it out.

Still, I would love to see a score of 60

[ioaz]

Ok, I think i'm an idiot, I ran the whole test. Were you guys just running the disk test? If that's the case I get about a 22.

[Apfhex]

So bigger numbers are better I take it.

Using Xbench 1.2 and doing ONLY the disk test, my PowerBook has a 7200 RPM drive which scored 31.31, and my external firewire drive, also 7200 RPM, scored 27.32. Is that right? What other factors determine this score? I had the idea I could do something similar if I got a Mac mini, but maybe it wouldn't be as noticeable?

[Kyros]

Someone needs to test this with a 10K rpm drive.

[moonmonkey]

Originally Posted by ioaz

Ok, I think i'm an idiot, I ran the whole test. Were you guys just running the disk test? If that's the case I get about a 22.

Yes, just the disk test

[mduell]

Originally Posted by Apfhex

So bigger numbers are better I take it.

Using Xbench 1.2 and doing ONLY the disk test, my PowerBook has a 7200 RPM drive which scored 31.31, and my external firewire drive, also 7200 RPM, scored 27.32. Is that right? What other factors determine this score? I had the idea I could do something similar if I got a Mac mini, but maybe it wouldn't be as noticeable?

The chipset that your external disk enclosure uses will affect the result, as will the model/speed of your external disk.

[moonmonkey]

Originally Posted by Kyros

Someone needs to test this with a 10K rpm drive.

I think the bottleneck is the Firewire 400 so it probably wouldn't make much of a difference.

Can you get 10,000 RPM drives?

[Zim]

Originally Posted by Simon

Much too complicated.

Don't reboot. Just open /Applications/Utilities/DiskUtility, select your mini's internal HD and go to the 'Restore' tab. As a source select the internal HD, as the destination select your new external FW disk. Have the utility format the disk and restore in one step. Go to the system prefs and select the external disk. Done. All preserved. It's as simple as that.

Simple yes... and slower than transcribing the bits by hand . never understood
why the Apple tools were so bloody slow....

(just got a new 200G drive and using this method to copy the contents of my 50G drive to it....30
minutes elapsed and only 25% done....) (powermac g4/800 to external FW400 drive, 10.4.2)

Mike

[osxrules]

All your tests seem pretty low - maybe it's because I'm using XBench 1.1.3. I only have a 1.25GHz Mini and my HD failed not long after getting it so I was forced to get an external firewire solution. Now my Mini came with one of the 5400rpm Seagates and my performance then with only the disk test was around 59:

Results 59.35
System Info
Xbench Version 1.1.3
System Version 10.4.2 (8C46)
Physical RAM 1024 MB
Model PowerMac10,1
Processor PowerPC G4 @ 1.25 GHz
L1 Cache 32K (instruction), 32K (data)
L2 Cache 512K @ 1.25 GHz
Bus Frequency 167 MHz
Video Card ATY,RV280
Drive Type ST940110A
Disk Test 59.35
Sequential 52.20
Uncached Write 94.11 39.23 MB/sec [4K blocks]
Uncached Write 71.49 29.28 MB/sec [256K blocks]
Uncached Read 25.32 4.01 MB/sec [4K blocks]
Uncached Read 79.86 32.27 MB/sec [256K blocks]
Random 68.76
Uncached Write 64.20 0.96 MB/sec [4K blocks]
Uncached Write 68.49 15.45 MB/sec [256K blocks]
Uncached Read 68.29 0.45 MB/sec [4K blocks]
Uncached Read 74.88 15.41 MB/sec [256K blocks]

For my external firewire case, I got a 7200 rpm Maxtor drive and I used SuperDuper to clone the internal drive. It took longer than I expected at about 1 hour for 30GB of stuff but it worked so I was happy. I then booted on my FW drive and ran XBench. I then got a value of 89:

Results 89.42
System Info
Xbench Version 1.1.3
System Version 10.4.2 (8C46)
Physical RAM 1024 MB
Model PowerMac10,1
Processor PowerPC G4 @ 1.25 GHz
L1 Cache 32K (instruction), 32K (data)
L2 Cache 512K @ 1.25 GHz
Bus Frequency 167 MHz
Video Card ATY,RV280
Drive Type Maxtor 6 Y080L0
Disk Test 89.42
Sequential 76.85
Uncached Write 94.93 39.57 MB/sec [4K blocks]
Uncached Write 68.58 28.08 MB/sec [256K blocks]
Uncached Read 61.30 9.70 MB/sec [4K blocks]
Uncached Read 94.15 38.04 MB/sec [256K blocks]
Random 106.92
Uncached Write 138.22 2.07 MB/sec [4K blocks]
Uncached Write 127.08 28.66 MB/sec [256K blocks]
Uncached Read 86.12 0.57 MB/sec [4K blocks]
Uncached Read 93.48 19.24 MB/sec [256K blocks]

Though the values differ, the performance increase is consistent with what others have found at 50% speed up. In real world terms, I didn't really notice much difference though. Perhaps I just didn't test it enough. It's not really an option for me though because my FW case has a fan that's always on so it's noisy. Plus, I prefer to use my external drive as more secure backup storage in case of a power surge or whatever. Plus, I prefer my backup drive to have more space than my everyday drive so I always have enough room.

BTW, my full test score comes out around 138. Maybe they changed the scoring in XBench 1.2.

[dougmilbourne]

Thanks for recommending this solution. I added a 200GB external firewire and performed the drive image. I haven't ran XBench on the new setup but the system is much "snappier"

[exactopposite]

Originally Posted by moon monkey

I think the bottleneck is the Fire wire 400 so it probably wouldn't make much of a difference.

Can you get 10,000 RPM drives?

the western digital raptor drives are 10k rpm. unless they recently updated they only come in 36 gig and 74 gig capacities, but they are FAST

[exactopposite]

oops

[exactopposite]

oops. double post

[frdmfghtr]

Out of curiousity, has anybody ever tried this via USB 2.0? I ask simply because I already have a USB 2.0 HDD enclosure with a 40 GB WD 7200 rpm drive that I use for weekly backups. From what little I've looked, USB 2.0 and Firewire are close to the same speed, with Firewire having a slightly higher sustained throughput.

frdmfghtr

[mduell]

Originally Posted by frdmfghtr

Out of curiousity, has anybody ever tried this via USB 2.0? I ask simply because I already have a USB 2.0 HDD enclosure with a 40 GB WD 7200 rpm drive that I use for weekly backups. From what little I've looked, USB 2.0 and Firewire are close to the same speed, with Firewire having a slightly higher sustained throughput.

Either the USB chipsets or drivers used in Macs are exceptionally poor.
Barefeats benchmarked USB and FW and compared them to PCs running Windows. Stick with FW on the Mac (or even better, SATA!).

[osxrules]

Originally Posted by mduell

Either the USB chipsets or drivers used in Macs are exceptionally poor.
Barefeats benchmarked USB and FW and compared them to PCs running Windows. Stick with FW on the Mac (or even better, SATA!).

I actually get pretty poor transfer rates on my external FW hard drive. I reckon it's on average 10-15 MB/s.

I don't think you can use USB2 for booting because wouldn't the driver have to be loaded before the drive could be recognised? AFAIK firewire doesn't require drivers.

[mduell]

Originally Posted by osxrules

I don't think you can use USB2 for booting because wouldn't the driver have to be loaded before the drive could be recognised? AFAIK firewire doesn't require drivers.

Firewire and USB2 both require drivers. Apple includes a low-level driver for Firewire in the machine's firmware so you can boot of Firewire drives; many PCs (like mine) can do the same from USB drives.

[westrock]

Originally Posted by exactopposite

the western digital raptor drives are 10k rpm. unless they recently updated they only come in 36 gig and 74 gig capacities, but they are FAST

The only problem being that I don't see any external cases with an internal SATA connector and an external Firewire connector...a couple USB ones, but no 1394

I'm sure some will pop-up soon as IDE goes to the hospice.

[mduell]

Originally Posted by westrock

The only problem being that I don't see any external cases with an internal SATA connector and an external Firewire connector...a couple USB ones, but no 1394

I'm sure some will pop-up soon as IDE goes to the hospice.

It's 5.25" instead of 3.5", but this is SATA on the inside and 1394 on the outside.
There's not much of a disadvantage to using PATA drives for externals... the Firewire bus is slower than PATA anyway.

[tooki]

Originally Posted by moonmonkey

I think the bottleneck is the Firewire 400 so it probably wouldn't make much of a difference.

Can you get 10,000 RPM drives?

They introduced 10,000 RPM drives 5 years ago. Now, the top end is 15,000 RPM. But don't expect to buy one any time soon -- the 146GB Seagate Cheetah 15K.4 costs $1100-1600, and it comes in Fibre Channel, Ultra320 SCSI, or Serial Attached SCSI interfaces that have no way to connect to a Mac mini.

15K RPM drives are, for now, strictly high-end server drives, with exceptional reliability and performance, but priced accordingly. (Which is why redundant arrays of cheap disks, like Xserve RAID, are gaining ground.)

tooki

[dannyillusion]

Wow, this was a great suggestion.
Just got an external FW HD and installed the system on it.
I can't believe the difference, so much faster!!

Thanks,
D

[techtrucker]

Why just Rev 1 of the mini? Does the latest model have a faster internal HD?

[moonmonkey]

Originally Posted by techtrucker

Why just Rev 1 of the mini? Does the latest model have a faster internal HD?

Yes, but still not as fast as an external firewire.

[Simon]

Originally Posted by moonmonkey

but still not as fast as an external firewire.

If the external disk runs at 7200rpm and the ATA-FW bridge chip used in the external disk enclosure is good.

[dc45]

I tried this with my 1.25 MacMini & OS 10.3.9(1024 MB RAM).
The XBench disk test with the internal drive read 27.04 overall.
When I "restored" the system on a DiamondMax Plus 9 80G drive in a Prolific Firewire case, the disk test measured 50.31. Nothing like some of the scores above, but it definitely did feel faster.
Can someone explain to me the difference between a 2MB buffer(my drive) and an 8MB buffer on some others? Is this Diamond Max a slow drive?

[mduell]

Originally Posted by dc45

Can someone explain to me the difference between a 2MB buffer(my drive) and an 8MB buffer on some others? Is this Diamond Max a slow drive?

The 2MB cache only takes 0.04 seconds to empty (assuming the right data is in it), while the 8MB cache will last for a whopping 0.16 seconds.
But if the 8MB drives are getting better Xbench scores I'd guess that a portion of the Xbench tests can fit within the 8MB cache, so they're never actually hitting the disk.

[dc45]

How much would Tiger (as compared to Panther)be a factor in these speed tests?

[dc45]

How much would Tiger-as opposed to Panther-be a factor?