huge thank you

Whoever at empeg added the download feature in emplode, thank you, thank you, thank you.

Those of us with MKI's couldn't use DisplayServer to get the files off so this is a huge thing.

Thanks again.

John

I agree, been sort of a lifesaver for me, since the HDD died on my PC that contained all my MP3 files.

Yes, i had the same problem, i had to restore my HD from my MK2. Someone should write a sync program to sync your hd with your empeg, so if either dies, it will restore the other, and if you add a song to your HD, it will update the empeg.

I'm hacking on something like this in my copious free time. Eventually, I might even finish it.

What?! If there is any critical data that is in need of RAID 5 protection, it's an MP3 collection!

I love my HP Netserver! :)

MP3 collections would really be better suited to a RAID3 configuration.

agreed. maybe a raid 5 using IDE? i have 5 drives, all same type, 60g. Doesn't Promise Make a controler board for that? Anyone know the price?

Promise makes an IDE raid controller that is basically a software solution, it is called the FastTrak, it is not OS independant because it relies on the Windows Driver. It works but the performance isn't great.

Adaptec now makes, or at least did make for a time an IDE raid controller, not sure on the details except it was $550 plus.

I have used several 3ware Escalade IDE RAID controllers, they show up to the system as SCSI devices and have excellent performance for most things, storagereview.com has more information.

There are also hardware IDE RAID solutions, we use one where I work from raidweb.com, 6 60GB IDE drives in RAID 0 (they also do 3 and 5) that comes out as SCSI and connects to a convential SCSI controller.

IDE RAID is a great solution for many applications and is far too often overlooked by ignorant SCSI snobs who don't know anything except that someone told them once SCSI was the way to go. Tragedy really, as most IDE drives are the exact same as their SCSI counterparts internally, and the good IDE RAID solutions only have 1 drive per channel and everything switched past there, so performance really isn't an issue...

- jason "dreaming of an Enlight 8590 with 8 120GB Maxtor IDE hard drives on an 8 port Escalade RAID card"

for personal backups. yeah.

I have the promise FastTrack66, and it seems to work fine. I HAD 2 striped drives but now that one drive died, i'm going to maybe stripe 2 and mirror another 2. Using 4 60gig drives, i'll have 120gig of space that's mirrored. Unless someone has a better solution, i think i'll do that.

Thanks for your input on the Raid Cards. However, i thought the Promise FastTrack WAS hardware Raid. It shows up in Windows as a scsi controller. i've used it in winNT4.0, 98, and 2000.

You should mirror 2 drives, mirror another 2 drives and stripe those mirrored sets together if your controller will allow that. This is called RAID1+0 or RAID10 and offers better protection than striping and then mirroring, which is referred to as RAID0+1. I'll explain further why that is, if anyone is actually interested.

the web site says only 0+1 and not the other way around. I would love to konw more on why 1+0 is better.

RAID 5 versus RAID 0+1 really depends on how large an array total and how much space you are willing to use for parity... probably a discussion best suited for another message board...

The promise card will not work at all in raid without drivers, and does not work under say... linux. The performance numbers also bear this out, it is very much dependant on the host CPU to think for it. The escalande does work in linux and will work in windows without any kind of drivers, though of course it does work better with the drivers installed.

In the end, it doesn't really matter as long as whatever you are using works for you, however you choose to define "works".

I prefer the 3ware product for a number of reasons, not the least of which being linux support, open source drivers, and better performance.

Promise won't release their specs allowing someone to build a linux driver, so I won't buy any more promise technology even though I was a great fan of their cards when they first came out and sold quite a lot of them.

I try to do my best to vote with my dollar and I think that having support for your favorite OS is a good thing, especially in the light of reverse engineering potentially being interpreted as everything from an act of treason on down...

Ramble ramble, I'm going home to contemplate hawking the rio car I just got on ebay for twice what I bought it for...

Now that I reconsider, using only 4 drives makes my statement incorrect. They are both equally likely to fail. The advantage of RAID1+0 comes when you have more than 4 disks. I'll still explain why, though.

Let me just state for the record that a mirror (RAID1) can survive with only one of its submirrors, but a stripe (RAID0) must contain all of its parts.

Let's first imagine a RAID1+0 setup. Let's say that there are six drives total. There are three mirrors, each with two submirrors. Let's call the mirrors A, B, and C and their respective submirrors 1 and 2. So the drives are labelled A1, A2, B1, B2, C1, and C2. If A1 fails, the A mirror is still active. Let's call that the initial failure state. If we then lose A2, then the RAID fails because mirror A fails and the stripe needs both parts of the stripe. But if we lose B1, B2, C1, or C2 then the RAID set remains active, because the mirrors, while most are in a degraded state, are all still active. So once in its initial degraded state, the chances of one other failure taking down the drive are 1 in 5. After that, 1 in 2. After that, 1 in 1.

Let's now imagine a RAID0+1 setup. There are still six drives, but now we have two striped sets, each with three drives, that are mirrored together. Let's call the striped sets A and B and their respective internal drives 1, 2, and 3. So we have drives A1, A2, A3, B1, B2, and B3. If A1 fails, the A stripe is disabled. It no longer has the data necessary to function. If we now lose any drive in the B stripe, the B stripe goes down and the RAID ceases to function. So after its initial degraded state, the chances of another single drive failure causing the RAID to go offline are 3 in 5. After that, 3 in 4. After that, 1 in 1.

So it's a statstics game, but it's not even close. And the odds become even further separated as you add more drives.

I'm sure you just misread my post, but just to clarify, RAID5 has nothing to do with RAID1+0/RAID10. They are wildly different RAID algorithms. Again, RAID3 is likely to be the optimal solution for MP3 storage, as it's engineered for long reads of consecutive data. RAID5 is unlikely to be the best solution for anything, except a small bank account.

IDE RAID is a great solution for many applications and is far too often overlooked by ignorant SCSI snobs who don't know anything except that someone told them once SCSI was the way to go.

True. People forget that original meaning of I in RAID was 'inexpensive'. First RAID boxes used ESDI or even ST506 drives (usually with SCSI external interface).

Yep... Promise makes some RAID cards, but their entry level stuff can only do RAID 0, 1 or 0+1. So your 4 60 Gig drives will give you 120 Gigs of fault tolerant space.

Adaptec makes some 4(?) channel IDE RAID cards that can do RAID 5. Bigger investment though. I see 'em pop up on ebay every now and then.

Linux can also do RAID{0,1,5} in software; ya just need the drives themselves.

-ml

As can Solaris, FYI. Which is also available free-of-charge.

Pardon my zealot streak, but Solaris is NOT "free of charge", unless you just want to play around with it. If you want to put your small biz onto it, a paid license is required.

All of "my" machines belong to my business, and thus none of them may use Solaris unless I want to pay big bucks for nothing. Ditto for that stuff from Redmond.

Cheers

-ml

No. It's actually free on any computer up to 8 processors. From http://www.sun.com/solaris/binaries/:

Now you can use the Solaris[tm] 8 Operating Environment at home or at work -- without paying a license fee.

The Promise card works without any drivers whatsoever. You create your RAID volume using the utility embedded in it's BIOS. From there, the RAID volume appears as one large hard drive using standard BIOS calls (IOW, fdisk would work the way it normally does). The drivers are needed for performance reasons, no matter what the OS is. If your OS can access the hard disk through the BIOS then it will be able to talk to the RAID volume (example, Windows 98 in safe mode). Most modern operating systems bypass the BIOS because it's too slow, hence the need for drivers.

Okay, cool. Thanks.

Not that that means that you should switch or anything. Linux is often better on x86 hardware. It supports way more stuff. And some of the stuff Solaris supports is a little flaky. But if you find a combination of Intel hardware that works right under Solaris, it tends to be more stable, in my experience. Linux seems to have more random glitches than Solaris does. But, then, you have to spend a lot of time finding that right combination to begin with. So I'd use Linux most of the time on x86 machines anyway. Really, I'd tend to use a BSD, but between the two....

And, BTW, you can download Solaris right off the net. You don't even have to pay for shipment anymore.

Well.. me personally, I wasn't really thinking of switching, for obvious reasons.

   find /usr/src/linux/ -type f -exec grep mlord {} \; 

Cheers

-ml

Actually, I suppose that should have been:

find /usr/src/linux/ -type f -exec grep 'mlord\|k Lord' {} \;

There!

Yeah, I knew that, but it should really be:

find /usr/src/linux -type f -print | xargs grep mlord

So, is that enough Unix sniping for you?

You lost the trailing slash on /usr/src/linux/ (needed for symlinks).

Auugggh! You got me. I usually include the following `.', as well, for some broken finds that I've had to deal with.

How reliable is it? All I have experience with is high end SCSI array controllers (Compaq's Smart Array line), and a buy with NT's software RAID (shudder).

Right now I am using LVM with RiserFS on my server at home, and the data spanned with the LVM is backed up to tape.

In case anyone found this mini-thread about the freeness of Solaris interesting (which you probably didn't ), I just found out that Sun has EOL'd Solaris x86. So if you were interested in seeing it, you'd better grab a copy now.

Actually, no they didn't. (Not sure why Slashdot posted that without verifying it...). According to Sun's FAQ on Solaris 9 Early Access, it says:

8. Why are there no media or download options for the Intel version of Solaris 9 OE?

We are defering the productization and release of Solaris 9 OE for Intel IA-32. Sun[tm] will continue to ship existing and new versions of Solaris 8 OE for Intel. We have made this decision for business reasons.

I know there are beta versions out there, as I have talked to a customer running it on their Intel server.

Actually, I didn't see it on Slashdot. I saw it on the SE Toolkit site, which seemed like a good source, considering Cockcroft's attachment to Sun. I've looked some more, and there are a lot of articles about it, many quoting Graham Lovell saying that it is ``deferred''. This sounds a lot like the equivalent of the television networks' ``on hiatus''. Still, there is a lot of opposition, which Sun seems to be noticing. Maybe they'll bring it back and actually support more than 5% of basic Intel hardware.

Maybe they'll bring it back and actually support more than 5% of basic Intel hardware.

Don't hold your breath. When Sun purchased Interactive Unix from Kodak in early 90's it was lightyears ahead of SCO (e.g. it was one of official launch distros for AT&T SVR4, both 'plain' and SMP) and Sun itself didn't have anything ready for Intel. THey scewed it up completely within few months and then discontinued it, leaving Intel market to SCO. Thank God for Linus & RMS!

With days of expensive Unices gone, I expect Sun focusing on their high-end hardware that still has perceived advantage over Linux (on Intel, Alpha or PowerPC) farms. I don't think that marketing value of Solaris for Intel as entry-level teaser is worth the trouble of supporting it...