Re: Software VS Hardware Raid
> > My question: if this was hardware RAID 1... would this have
> > happened? Would the hardware RAID controller recognise the
> > problem, and only stop briefly, then try the second disk
> > automatically and transparently?
>
> I believe hardware RAID would have worked transparently here.
>
> With software RAID, you may have to make your /boot partition
> (with your kernels etc.) non-RAID and then when the first disk
> fails, manually swap the second one so that it is the primary
> master. I don't know if there's a way to get it to boot the
> second drive automatically.
There are two ways. One way makes lilo recognize the raid at bootup,
another way you put /boot on both disks, and whichever boots up first gets
to start the RAID (I'm no expert on RAID so feel free to correct me if
wrong).
>
> You might be able to do it if you boot from a bootloader (e.g.
> grub) on a stiffy disk and then get that to boot from the hard
> drives, but then you'll be relying on a stiffy disk :) And I
> don't know exactly if/how this would work anyway.
>
> If you could get it to work from a stiffy disk, you could use a
> flash ROM (e.g. a DiskOnChip) instead of the stiffy disk.
I fear adding another level of complexity to the whole thing. Imagine if
something ever goes wrong to some of this stuff... eg. settings. It would
become increasingly impossible to track down the exact cause of the error
or recover from it. There are those USB "flashable" roms that are
selling... they have from 8-256M, but its USB, and no one (yet) has made a
way to boot from them afaik.
> > Case 2)
> > I simulated errors by connecting a flaky IDE cable to one of
> > the drives. I was hoping the software RAID would either
> > compensate by doing most of it's reading from the good drive
> > (with a good cable) or labelling the flaky cable/drive as bad,
> > but instead it started slowing down, and writing to the array
> > was taking much longer and strange errors starting occurring
> > during writing.
> >
> > My question: would hardware raid have handled this situation
> > any better?
>
> I am not sure, but I think there's a good chance that hardware
> RAID would also have trouble with this. Maybe you should ask
> some hardware RAID companies or get them to lend you a
> controller for evaluation.
We're only going to be buying about 6-10 of these cards... not a 100. I
doubt the vendors would give us any for evaluation :-/
> > And as for Hardware IDE raid, which is better... Promise or
> > HighPoint? promise seems to be better supported in the
> [snip]
>
> You should also look at 3ware. http://www.3ware.com/
>
> I can't say which is better, because I've never used any of
> them, but 3ware has open source drivers in the kernel.
>
> From linux/Documentation/Configure.help:
>
> 3ware is the only hardware ATA-Raid product in Linux to date.
> This card is 2,4, or 8 channel master mode support only.
>
Only one slight nagging problem... read this:
Dear Valued 3ware Customer,
3ware is pleased to announce that due to overwhelming customer feedback,
it will continue full support, development and production for it's popular
Escalade products. The Escalade product family, first introduced in
December, 1999, was the first RAID controller for ATA disk drives to
utilize patented StorSwitchT technology. 3ware's innovative approach
provides very high performance RAID solutions using lower cost ATA disk
drives.
The Escalade 7000 product line will continue with 4 products. 7410, 7810,
7450, 7850. They are available in retail kits or bulk 10 paks.
Somehow the above doesn't inspire a lot of confidence in me that they are
going to be full heartedly continuing the products that I am interested in
(IDE RAID)... and I certainly don't want to buy something, only to find
out soon after that they've decided to not continue.
Reply to: