RAID disk replacement

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.storage (More info?)

Generally speaking, in a RAID setup, if a single drive fails it would seem
ideal to replace it with the exact same make/model drive if possible.
However, if this is not possible due to the age and lack of availability of
the failed drive, is it certain that the drive can replaced by a different
make/model one, so long as both the speed and the capacity are equal to or
greater than the original?

i.e. I could replace a 7,200 RPM 36GB drive with a 10,000 RPM 36GB drive
with no (apparent) problem? I wouldn't expect any performance gains from
having a single drive faster than the others...data integrity is the single
concern for my question.

I'm currently faced with this problem, and I'm getting some conflicting
info about the exact criteria required of the replacement drive in a RAID
system

That being the case, I'd especially like to hear from someone(s) who has
done this successfully in actual practice,...thanks!

_______________________________________________________________________________
Posted Via Uncensored-News.Com - Accounts Starting At $6.95 - http://www.uncensored-news.com
<><><><><><><> The Worlds Uncensored News Source <><><><><><><><>

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.storage (More info?)

Jake Lee wrote:

> Generally speaking, in a RAID setup, if a single drive fails it would seem
> ideal to replace it with the exact same make/model drive if possible.
> However, if this is not possible due to the age and lack of availability
> of the failed drive, is it certain that the drive can replaced by a
> different make/model one, so long as both the speed and the capacity are
> equal to or greater than the original?

I wouldn't go so far as to say it was _certain_ because sure as I do you're
going to find that you have some pathological combination of hardware and
software that won't allow it.

That said, I've never had any trouble replacing a member of a RAID
(hardware, software, ATA, SCSI, ESDI, or MFM) with a larger drive--the only
real rules I've found are that the new drive has to have the correct
interface, has to be compatible with whatever software and hardware you're
using, and has to have the same or greater capacity than the one it
replaces.

> i.e. I could replace a 7,200 RPM 36GB drive with a 10,000 RPM 36GB drive
> with no (apparent) problem? I wouldn't expect any performance gains from
> having a single drive faster than the others...data integrity is the
> single concern for my question.
>
> I'm currently faced with this problem, and I'm getting some conflicting
> info about the exact criteria required of the replacement drive in a RAID
> system
>
> That being the case, I'd especially like to hear from someone(s) who has
> done this successfully in actual practice,...thanks!
>
>
_______________________________________________________________________________
> Posted Via Uncensored-News.Com - Accounts Starting At $6.95 -
> http://www.uncensored-news.com
> <><><><><><><> The Worlds Uncensored News Source
> <><><><><><><><>

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.storage (More info?)

In article <c5cill02795@news4.newsguy.com>,
jclarke@nospam.invalid says...
> Jake Lee wrote:
>
> > Generally speaking, in a RAID setup, if a single drive fails it would seem
> > ideal to replace it with the exact same make/model drive if possible.
> > However, if this is not possible due to the age and lack of availability
> > of the failed drive, is it certain that the drive can replaced by a
> > different make/model one, so long as both the speed and the capacity are
> > equal to or greater than the original?
>
> I wouldn't go so far as to say it was _certain_ because sure as I do you're
> going to find that you have some pathological combination of hardware and
> software that won't allow it.
>
> That said, I've never had any trouble replacing a member of a RAID
> (hardware, software, ATA, SCSI, ESDI, or MFM) with a larger drive--the only
> real rules I've found are that the new drive has to have the correct
> interface, has to be compatible with whatever software and hardware you're
> using, and has to have the same or greater capacity than the one it
> replaces.
>

That matches exactly with what I've found. Also makes
for a slow-n-steady way to end up with a larger array
(once you've replaced the last small drive with a bigger
drive, backup-and-reinstall to take advantage of the
larger drives).

The other rule-of-thumb that I've heard when buying
initial drives for an array is to match the make-model
as close as possible. But make sure all of the drives
come from different manufacturing batches. (e.g. buy 1
here, 1 there, 1 from somewhere else, etc.) The idea
being that if there is a fatal flaw in a particular
batch, you won't have multiple drives fail at the same
time.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.storage (More info?)

"Toshi1873" <toshi1873@nowhere.com> wrote in message
news:MPG.1ae525d17674d1af989845@news-50.giganews.com...
> In article <c5cill02795@news4.newsguy.com>,
> jclarke@nospam.invalid says...
> > Jake Lee wrote:
> >
> > > Generally speaking, in a RAID setup, if a single drive fails it would
seem
> > > ideal to replace it with the exact same make/model drive if possible.
> > > However, if this is not possible due to the age and lack of
availability
> > > of the failed drive, is it certain that the drive can replaced by a
> > > different make/model one, so long as both the speed and the capacity
are
> > > equal to or greater than the original?
> >
> > I wouldn't go so far as to say it was _certain_ because sure as I do
you're
> > going to find that you have some pathological combination of hardware
and
> > software that won't allow it.
> >
> > That said, I've never had any trouble replacing a member of a RAID
> > (hardware, software, ATA, SCSI, ESDI, or MFM) with a larger drive--the
only
> > real rules I've found are that the new drive has to have the correct
> > interface, has to be compatible with whatever software and hardware
you're
> > using, and has to have the same or greater capacity than the one it
> > replaces.
> >
>
> That matches exactly with what I've found. Also makes
> for a slow-n-steady way to end up with a larger array
> (once you've replaced the last small drive with a bigger
> drive, backup-and-reinstall to take advantage of the
> larger drives).

Well in RAID 0 or RAID 1 then you are limited to the size of the smallest
drive.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.storage (More info?)

Previously Jake Lee <none@nope.com> wrote:
> Generally speaking, in a RAID setup, if a single drive fails it would seem
> ideal to replace it with the exact same make/model drive if possible.
> However, if this is not possible due to the age and lack of availability of
> the failed drive, is it certain that the drive can replaced by a different
> make/model one, so long as both the speed and the capacity are equal to or
> greater than the original?

Yes. But take care that it is really the same capacity or larger,
i.e. has at least the same number of sectors.

> i.e. I could replace a 7,200 RPM 36GB drive with a 10,000 RPM 36GB drive
> with no (apparent) problem? I wouldn't expect any performance gains from
> having a single drive faster than the others...data integrity is the single
> concern for my question.

As long as the new disk is really larger. Exact disk sizes may differ
between models/manufacturers by some tens of MBs or more. If the
new disk is just one sector smaller, things may not work.

> I'm currently faced with this problem, and I'm getting some conflicting
> info about the exact criteria required of the replacement drive in a RAID
> system

> That being the case, I'd especially like to hear from someone(s) who has
> done this successfully in actual practice,...thanks!

Only for software-RAID under Linux. The only limitation is the drive
size there.

Arno
--
For email address: lastname AT tik DOT ee DOT ethz DOT ch
GnuPG: ID:1E25338F FP:0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
"The more corrupt the state, the more numerous the laws" - Tacitus