Hmm. Not Seagate's fault.

[mdlbear]

The new disk, in an external USB/eSATA enclosure, wouldn't stay up more than a couple of minutes at a time. And it wasn't spinning, down, either: just locking up. OK, that's not a hardware problem; it's probably the kernel not dealing properly with something in SATA II. Installed 2.6.22 from backports, and it seems to have worked. USB was horribly slow, and it doesn't seem to handle hotplugging an eSATA drive, but after a reboot it seems perfectly happy. The same kernel in the fileserver has been up ever since I installed it, and my Ubuntu-based recording studio hasn't had any problems.

So I'll have a couple of 400GB SATA drives free pretty soon. Three once I upgrade the mirror drive, but I want to wait for the price to come down a little. RAID box?

1/14 I may have spoken too soon -- there may also be some kind of controller issue. Or something. I'm dropping back to IDE on the fileserver.

Comments

I know what teh problem is

[trogula.livejournal.com]

Power. Especially with a drive that large.

The larger the external disk, the more power it needs. Essentially, I've found that anything over 200GB requires an external power supply for the enclosure, even if your USB is power-providing (all are, I believe).

I've seen exactly this type of behavior on Windows, OS X, and Linux several times - all were solved by an external power supply.

Please ignore if you already have an external power supply - this could then be something else.

Re: I know what teh problem is - no, you don't

[mdlbear]

It's a 3.5" drive, and the enclosure's power supply has 2A a 5V and 2A at 12V, and is spec'd for 500GB drives. The problem also occurs with internal drives and is fixed completely with a 2.6.22 kernel.

I belatedly remembered having the same problem on my recording box, and how UbuntuStudio was the only distro that worked properly. It has 2.6.22.

Re: I know what teh problem is - no, you don't

[trogula.livejournal.com]

No worries. Like I said - just sharing what I've seen a number of times.

Re: I know what teh problem is - no, you don't

[mdlbear]

I've seen power problems on external enclosures, too. I've noticed that they come with bigger power supplies these days. But these problems are across the board; I'm inclined to suspect a race condition in the driver that shows up with faster drives.

In any case, no longer a problem, now that I know about it. And as long as I don't try to install Etch from scratch on one of the Seagate drives.

Re: I know what teh problem is - no, you don't

[mdlbear]

I spoke too soon; evidently it is still a problem. Most likely having to do with new drives, old controllers, and flaky drivers. I am grumpy about it, but I can install an IDE drive and make it stop.