RAID newbie...can I have several partitions on a RAID 1 array?

[Odie Ferrous]

Which isn't worth a damn if the axle broke.

By the same token, if a motherboard or RAID controller failed, it would
also have the immediate effect of zero redundancy. There again, though,
the axle could be repaired, inasmuch as the motherboard or RAID
controller could.

Why bother posting at all, Folkert? You get it wrong, with practically
every post you make.

Perhaps you should consider early retirement from usenet.



Odie

 

[willbill]

I'm thinking of setting up a RAID 1 array with two 250GB Drives (7K250
SATAII )

This is my first venture into RAID so I have a few newbie questions...

Can I split the drive into several partitions or does it have to remain as a
single partition?

the "drive" (raid array) can be split into
several partitions

If I can, will most software partition managers (PQ Partition Magic,
Acronis, Ghost) cope with RAID or do I need a special partition manager? Can
you recommend one?

odds are you are using Win XP or 2000 or Linux

do the final partitioning in the OS "Computer Management"
(XP: >start>programs>administrative tools>computer management)

fwiw, my Acronis True Image 9 (a backup program) does
not see raid 1 accurately (when i boot from the TI9 CD),
so i'd be cautious about any separate partition manager

As far as my software is concerned I presume they ignore the RAID aspect and
just see it as a normal single drive?

correct


Just in case your were wondering... I am aware that RAID 1 is a mirrored
array. I'll be using the Sil 3114 RAID controller on my ASUS A8N-SLI Premium
Mobo.

this is a late response, so i'm assuming
that you've already battled thru this

anyway, i have the Sil 3114 controller on
my Tyan S2875 mobo

the 1st step is to enter into the bios setup
(press the delete key when the boot messages
start to show up on the monitor), and find
the entry that gives options for the
SATA drive connectors

my own two options are: ultra or raid
(it doesn't get specific about raid options,
that's in the 3114 s/w (below))

save any changes, reboot, and during
the boot hit the pause button when
the 3114 display messages appear

there will be a keyboard key that you
have to push to enter into the 3114
s/w to define the raid array.
it takes very little time to execute
to completion for a 2 disk raid 1
(no striping needed for raid 1)

reboot again. this time all the
way into your OS, and do the final
partitioning there

bill

 

[Folkert Rienstra]

If you want an answer, do away with the clever insults....

If that will shut you up ....

 

[Gerhard Fiedler]

Everybody knows you should always carry a spare (I mean redundant)
axle for "true redundancy."

Every spare is by definition redundant, else it would be needed and
wouldn't be a spare. Of course there are other forms of redundancy, too.

Gerhard

 

[Curious George]

Every spare is by definition redundant, else it would be needed and
wouldn't be a spare. Of course there are other forms of redundancy, too.

Gerhard

You've missed the point.

 

[Curious George]

If you want an answer, do away with the clever insults....

Arno

I haven't asked you a question.

 

[Arno Wagner]

Curious George wrote:

Every spare is by definition redundant, else it would be needed and
wouldn't be a spare. Of course there are other forms of redundancy, too.

Quite obvious, I would think. Why some people believe it has
to be automatic failover to be redundancy is beyond me.
Preconceptions do not make truths....

Arno

 

[Curious George]

Quite obvious, I would think. Why some people believe it has
to be automatic failover to be redundancy is beyond me.

That's because you're mangling terminology and the arguments and
concepts at hand.

This attempt to convey this subthread singularly as a debate as to
whether or not a "spare" is "redundant" in a general sense is a red
herring.

Preconceptions do not make truths....

Indeed.

 

[Folkert Rienstra]

Every spare is by definition redundant, else it would be needed and
wouldn't be a spare.

Impeccable reasoning as always, Fiddler.

So in your reasoning you should just take one disk out of a RAID array,
otherwise it isn't redundant, right?

Of course there are other forms of redundancy, too.

No shit.

 

[Gerhard Fiedler]

Impeccable reasoning as always, Fiddler.

No shit

So in your reasoning you should just take one disk out of a RAID array,
otherwise it isn't redundant, right?

Wrong, of course. Just because a cold spare is redundant doesn't mean that
something that isn't a cold spare is not redundant. Logic 101 (or even less
than 101

BTW, re "true redundancy", that's a silly expression IMO. Either there is
or there isn't redundancy. There's no "false redundancy"...

Gerhard

 

[Arno Wagner]

Folkert Rienstra wrote:

No shit

Wrong, of course. Just because a cold spare is redundant doesn't mean that
something that isn't a cold spare is not redundant. Logic 101 (or even less
than 101

BTW, re "true redundancy", that's a silly expression IMO. Either there is
or there isn't redundancy. There's no "false redundancy"...

I used it to stress that it was not "partial redundancy". It
seems a number of people have problems with the term. Maybe "full
redundancy" or "redundancy for every component" would have been
clearer.

Arno

 

[Gerhard Fiedler]

I used it to stress that it was not "partial redundancy". It seems a
number of people have problems with the term. Maybe "full redundancy" or
"redundancy for every component" would have been clearer.

I don't think full redundancy in this sense is even possible on a system
level. There probably will always be a single point of failure in the part
that controls which of the redundant subsystems is currently active. (In
the case of a simple RAID1 array it's the controller and everything up in
the chain, in the case of cold spares it's the system between keyboard and
chair

In the end, it's not about redundancy in itself. The various forms of
redundancy are only part of the measures to reduce downtime and/or failure.

Gerhard

 

[Arno Wagner]

Arno Wagner wrote:

I don't think full redundancy in this sense is even possible on a system
level. There probably will always be a single point of failure in the part
that controls which of the redundant subsystems is currently active. (In
the case of a simple RAID1 array it's the controller and everything up in
the chain, in the case of cold spares it's the system between keyboard and
chair

True, of course ;-)

In the end, it's not about redundancy in itself. The various forms of
redundancy are only part of the measures to reduce downtime and/or failure.

And to make recovery cheaper/less effort on average.

Arno

 

[Ian R]

Hi

I'm thinking of setting up a RAID 1 array with two 250GB Drives (7K250
SATAII )

This is my first venture into RAID so I have a few newbie questions...

Can I split the drive into several partitions or does it have to remain as
a single partition?

If I can, will most software partition managers (PQ Partition Magic,
Acronis, Ghost) cope with RAID or do I need a special partition manager?
Can you recommend one?

As far as my software is concerned I presume they ignore the RAID aspect
and just see it as a normal single drive?

Just in case your were wondering... I am aware that RAID 1 is a mirrored
array. I'll be using the Sil 3114 RAID controller on my ASUS A8N-SLI
Premium Mobo.

Thanks for your time and any info.

Ian.

Gosh!

Its amazing how asking a simple question can (de?)generate into a thread of
so much mud slinging and insults!

Well, thanks to all of you for replying with helpful info.

Just wanted to let you know that I have now successfully set up my RAID 1
array.

My main reason for doing this is to provide an extra degree of safety when
storing other peoples data on my system. I do realise its not as secure or
reliable as having an off system back up. But at least until thats done
(i.e. to USB HDD or DVD) in this case a mirrored array has got to be better
than a single drive.

Thanks again.

Ian I^)

 

[Folkert Rienstra]

No shit


Wrong, of course. Just because a cold spare is redundant doesn't mean that
something that isn't a cold spare is not redundant. Logic 101 (or even less
than 101

BTW, re "true redundancy", that's a silly expression IMO. Either there is
or there isn't redundancy.

Pity it is you who keeps speaking of "various" forms of redundancy.

There's no "false redundancy"...

There has to be if there is "true" redundancy, as part of your "various" .

 

[Folkert Rienstra]

Gosh!

Who's he?

Its amazing how asking a simple question

That's what troll questions usually are.

can (de?)generate into a thread of so much mud slinging and insults!

Don't want to troll, don't ask a simple question.

 

[Curious George]

Gosh!

Its amazing how asking a simple question can (de?)generate into a thread of
so much mud slinging and insults!

Simple question? Gosh, I see a multi-part "educate me as much as
possible" type question.

Well, thanks to all of you for replying with helpful info.

Just wanted to let you know that I have now successfully set up my RAID 1
array.

If you committed to RAID1 AFTER this thread, and have no questions or
concerns about it, you either didn't read it carefully or were
trolling, as Folkert suggested.

My main reason for doing this is to provide an extra degree of safety when
storing other peoples data on my system.

Of course you needed to explain that. But it will provide no "extra
degree of safety" or reliability. It can only provide uptime/
availability through _some_ kinds of disk problems - and not
necessarily the most common kinds.

I do realise its not as secure or
reliable as having an off system back up.

Apples and oranges. Fault tolerance has nothing to do with backup.
They don't even belong in the same sentence. They exist for very
different reasons and therefore require totally different plans and
goals.

"Backup" is one of the starting points of the discussion of the wisdom
of holding a spare controller. Another point not mentioned is that
you're using an embedded controller. You would be stuck with either a
second motherboard or tracking down a compatible spare (the SiI 3114
on PCI is not widely available). If you have sound backup system and
are expecting/requiring increased reliability rather than availability
necessarily - you are barking up the wrong tree both with "raid on the
cheap" and holding a cold-spare controller.

But at least until thats done
(i.e. to USB HDD or DVD)

Unless you have VERY little data, a lousy idea IMHO.

CD-DVD is also especially poor if there are frequent file changes. It
turns into lots of manual media changes and backup jobs.

USB HDD is especially poor if there are lots of small files or very
large backups. It takes forever and drives are often prone to
overheating.

There is also fragility and shelf-life of the media to consider.

in this case a mirrored array has got to be better
than a single drive.

It isn't. An array is mathematically less reliable than a single
drive. Furthermore an array of that caliber cannot possibly meet your
apparent expectations.

 

[Ian R]

BIG SNIP>


Simple question? Gosh, I see a multi-part "educate me as much as
possible" type question.


If you committed to RAID1 AFTER this thread, and have no questions or
concerns about it, you either didn't read it carefully or were
trolling, as Folkert suggested.


Of course you needed to explain that. But it will provide no "extra
degree of safety" or reliability. It can only provide uptime/
availability through _some_ kinds of disk problems - and not
necessarily the most common kinds.


Apples and oranges. Fault tolerance has nothing to do with backup.
They don't even belong in the same sentence. They exist for very
different reasons and therefore require totally different plans and
goals.

"Backup" is one of the starting points of the discussion of the wisdom
of holding a spare controller. Another point not mentioned is that
you're using an embedded controller. You would be stuck with either a
second motherboard or tracking down a compatible spare (the SiI 3114
on PCI is not widely available). If you have sound backup system and
are expecting/requiring increased reliability rather than availability
necessarily - you are barking up the wrong tree both with "raid on the
cheap" and holding a cold-spare controller.


Unless you have VERY little data, a lousy idea IMHO.

CD-DVD is also especially poor if there are frequent file changes. It
turns into lots of manual media changes and backup jobs.

USB HDD is especially poor if there are lots of small files or very
large backups. It takes forever and drives are often prone to
overheating.

There is also fragility and shelf-life of the media to consider.


It isn't. An array is mathematically less reliable than a single
drive. Furthermore an array of that caliber cannot possibly meet your
apparent expectations.

Hello Curious George

Thanks for replying.

Well for the record I'm not a troll.

From what youve said I think youre basically saying that RAID 1 is pretty
much a waste of time unless Ive misunderstood something (quite possible).

I can see that if my motherboard died and I wanted to access the data on my
RAID array then I would have to replace the mobo with the same one or at
least another that had the same controller.

If one of the drives in the array failed then I cant see how that would be
worse than having a single stand alone drive. At least my data would be safe
on the mirrored drive How come you dont see any benefit in having a
mirrored array?

I realise that DVD backup is vulnerable and slow as is USB HDD but given the
amount of data that many of us are having to backup this is a common method.

However being rather impatient and tired of umpteen disc changes to DVD I
bought a couple of mobile HDD racks and a couple of 250Gb HDD's. So I can
now backup to SATA HDD and swap the drives as needed - so much faster and
easier. And much cheaper than any alternative I can see.

YMMV

Cheers.

Ian

 

[Gerhard Fiedler]

Pity it is you who keeps speaking of "various" forms of redundancy.


There has to be if there is "true" redundancy, as part of your "various" .

"Various forms" doesn't imply that "true" and "false" are "forms" of
redundancy. Try designing several redundant systems to real-world
requirements, and you'll discover (possibly) some of the various forms of
redundancy.

Cold spares, running subsystems in parallel, having spare subsystems
available for automatically putting them in operation but not actually
running them in parallel, having completely different subsystems available
(like battery supply and mains supply or several different navigation
systems on a plane), and so on.

There are many forms of redundancy. None of which are "true" or "false"
redundancy... they are simply redundancy.

Gerhard

 

[Arno Wagner]

Folkert Rienstra wrote:

"Various forms" doesn't imply that "true" and "false" are "forms" of
redundancy. Try designing several redundant systems to real-world
requirements, and you'll discover (possibly) some of the various forms of
redundancy.

Cold spares, running subsystems in parallel, having spare subsystems
available for automatically putting them in operation but not actually
running them in parallel, having completely different subsystems available
(like battery supply and mains supply or several different navigation
systems on a plane), and so on.

There are many forms of redundancy. None of which are "true" or "false"
redundancy... they are simply redundancy.

As I explained, was using "true" to amplify that it is not partial
redundancy, i.e. there are no single points of failure left.

Arno