Guide to Create a Portable External FireWire Drive for Mac OS 9 w/ Partitioning

[IIO]

This is the absolutely WORST external case I've ever tried!

If you set it to RAID 0, and one of the two drives fails: you lose EVERYTHING on BOTH drives.

that is probably not the fault of the enclosure, because all other controllers will give you the same result.

if you need redundancy, use redundancy.

If you set it to RAID 1, so that if one of the two drives fails, you still have access to your data, you find that the data on the good drive is UNAVAILABLE until you remove the good drive and put it in a good external case.

that is also to be expected. when one of two mirrored drives is not available, you must replace it with new one in order to start the automatic rebuild. that is how redundancy (and therefore data integrity) is realized.

operating half a raid composite is like trying to drive a bicycle with only one wheel.

[mrhappy]

like trying to drive a bicycle with only one wheel.
[/quote]

I tried that once... that's why I walk with a limp!

[DieHard]

that is also to be expected. when one of two mirrored drives is not available, you must replace it with new one in order to start the automatic rebuild. that is how redundancy (and therefore data integrity) is realized.

operating half a raid composite is like trying to drive a bicycle with only one wheel.

NOT so... every RAID 1 I have worked with in 20 years, both internal and via a case still functions and mounts when degraded

If you set it to RAID 0, and one of the two drives fails: you lose EVERYTHING on BOTH drives.

Well that's normal with any "performance" stripping, RAID 0 does this by nature, you can't blame the case for implementing the specification

If you set it to RAID 1, so that if one of the two drives fails, you still have access to your data, you find that the data on the good drive is UNAVAILABLE until you remove the good drive and put it in a good external case.

OK, now that DOES suck, I have never had a drive fail in one of these as I standardized on the OWC Dual mini, and have only use the Maximus over the last 2 years; I have 3 Dual OWC Minis and 2 Maximus.... never had a failure on the maximus... you can remove a single drive from the OWC and it mounts as is readable internally on a mac... I will look into this Maximus thing, which if you are right SUCKS, this assumes you always have a "good" spare on the shelf.
So, again, to clarify, the External is NOT functional with a single drive failure ?!  It goes into "your screwed... use me after you replace the bad drive and I re-mirror" mode... this is NOT as expected.  The case should be functional with a failed drive, like I stated, a RAID 1 system should still be usable when "degraded" 

The Guardian Maximus Mini is NOT a good external case. It is a stinker.

Hmmm... you maybe correct, I will be testing this when I get some time

The safety it promises is totally missing. Not only can't you access your data, because it doesn't mount

Wow... not good

it DOES show up as being available to format. So after losing a bad drive, it becomes very easy to accidentally erase the good one.

Do you mean when you attach the good drive on an internal mac controller ?

[IIO]

NOT so... every RAID 1 I have worked with in 20 years, both internal and via a case still functions and mounts when degraded

might be that there are cases which do that, but it is not what i would exspect.

unlike with duplexing in n enterprise situation, a raid controller such as you can find in enclosures typicallly writes data only once - but to both drives (mirroring)

now if one drive is missing or broken, why should the other one mount when it is not possible to write to it?

you might want to switch the controller to jbod to access your data, that should always work. i would still remove the broken drive though...

[DieHard]

now if one drive is missing or broken, why should the other one mount when it is not possible to write to it?

I Think we are getting confused here, as far as a RAID 1 setup when a single drive fails the RAID goes into a "degraded" state, the Good drive should mount without issue, the Bad obviously will give all sorts of problems

you might want to switch the controller to jbod to access your data, that should always work. I would still remove the broken drive though...

OK, as far as changing settings from RAID to JBOD (Just a bunch of disks), I do not recommend this, some controllers will zero a drive is you mess with the RAID info, so DO NOT EVER switch a drive's mode if you want the data intact, most Intel Embedded RAID setups will simply remove the "RAID definition" and NOT reformat the drive, so that is a rare exception to the rule; Again, most mirrored (RAID 1) implementations are meant to Mount and work and simply message the user that a drive has failed...

Back to the Ext case thing, since that is the context we were talking about...

NEVER change dip switches or redefine RAID via dials or switches (if you care about the data) since many external boxes that do there own hardware RAID implementation will "nuke" the partition table and basically reformat the drive when changing the RAID definition 0,1,JBOD(no raid).  I can tell you that drives removed from most Firewire RAID ext cases/products I have used over the years as far a OWC, Lacie, and other mac vendors are totally readable if you put the drive on an internal controller. 

Some NAS products, like some of the fucking western digital boxes (and seagate boxes) utilize "encryption" on their circuit boards to the point that all drives show up as "unformatted" if you take them out of the case... a real bummer, to the point you have to get (from china) the exact revision on the Backplane/controller board is you get a fried case...which happens more than you think;  even cases that look identical and may be the exact model will not read the drives, if the board rev in different.... so, I HATE Propriety NAS boxes for this reason, you better have an extra case (exact model) around if your NAS logic board fries

[IIO]

your right, i´ve read up the specs and normally only one drive should keep operating normal when the other one isnt readable.

maybe the issue here is that that other drive hangs during mount or something.

removing the broken one should fix that.

removing the good one contains the risk that when you later put it in gain, it might be "restored" from the replaced drive.

[DieHard]

removing the broken one should fix that.

agreed   I was thinking the same thing, maybe the defective drive has a short or something that is stopping "normal" operation, with the dab drive removed, it should "limp" along... we need to confirm if he left the bad drive in the case

[mrhappy]

we need to confirm if he left the bad drive in the case

Well he never actually said he had a drive failure did he? I thought he was just complaining about how the case/raid functioned?? Hopefully he'll stop back with more info.

I bought two of those and have been running some fairly heavy PT sessions with them and they've been great for that. Was about to order a few more but holding off till the dust settles.

I never used raid before but so far I'm liking it! Have a question tho..

The Whole box shows up as a single drive (using Raid 1)... is that how it works? I figured that I'd be able to see them individually but that isn't the case. The lights on the display confirm that all is well so I'm believing it ... for now at least!