How does RAID 1 work in restore situation - Briefly.

  • Thread starter Thread starter ThomasAJ
  • Start date Start date
T

ThomasAJ

Guest
I know Raid 1 uses 2 identical disks and stores the data identically on each
one.
I also only want to talk about hardware raid. I have a very simple W2003
setup with 5 users on Terminal Services.

I want to upgrade my 3YO IBM X206 with a new server that has hardware raid
and 2 drives. The new server has cold-swap drives feature. (I cannot get any
technical person from IBM to talk to me - go figure)

So what happens when a drive fails?
Does the W2003 know about? Does it issue a message? Does it stop working
altogether until whatever is done?
Should I purchase a 3rd IDENTICAL drive so when one fails it can take over?
Does it takeover automatically? or do I replace the fallen over one with it?

OK you see I have no experience at the recovery end. I have spent a lot of
hours scouring the IBM site in order to find out how raid works from
beggining to end. No luck.

--
Regards
Tom
 
Re: How does RAID 1 work in restore situation - Briefly.


"ThomasAJ" <ThomasAJ@discussions.microsoft.com> wrote in message
news:FBBEBA70-5D7F-4403-BA0A-47DFEC2EE53D@microsoft.com...
>I know Raid 1 uses 2 identical disks and stores the data identically on
>each
> one.
> I also only want to talk about hardware raid. I have a very simple W2003
> setup with 5 users on Terminal Services.
>
> I want to upgrade my 3YO IBM X206 with a new server that has hardware raid
> and 2 drives. The new server has cold-swap drives feature. (I cannot get
> any
> technical person from IBM to talk to me - go figure)
>
> So what happens when a drive fails?
> Does the W2003 know about? Does it issue a message? Does it stop working
> altogether until whatever is done?
> Should I purchase a 3rd IDENTICAL drive so when one fails it can take
> over?
> Does it takeover automatically? or do I replace the fallen over one with
> it?
>
> OK you see I have no experience at the recovery end. I have spent a lot of
> hours scouring the IBM site in order to find out how raid works from
> beggining to end. No luck.
>
> --
> Regards
> Tom


You're talking about a disk mirror. When one of your disks fails
then the system generates an error message but it continues
operating normally. This gives you the opportunity to shut down
the system, replace the failed disk and rebuild the mirror. There
is no automatic failover in a RAID1 configuration.

To gain confidence, you should go through a practice run with
you mirror. Unplug the data cable and watch what happens (but
switch the system off before you plug it back in again!). See
if you can boot the machine with one or the other mirror disks.
 
Re: How does RAID 1 work in restore situation - Briefly.

When you say "unplug the data cable" can you please elaborate on that.

Also should I purchase an identical 3rd. drive. Bear in mind that failure
may occur 4 years down the track and the original drive may not be available
and if it is then shipping may take a while during which time there will be
no RAID 1.

--
Regards
Tom


"Pegasus (MVP)" wrote:

>
> "ThomasAJ" <ThomasAJ@discussions.microsoft.com> wrote in message
> news:FBBEBA70-5D7F-4403-BA0A-47DFEC2EE53D@microsoft.com...
> >I know Raid 1 uses 2 identical disks and stores the data identically on
> >each
> > one.
> > I also only want to talk about hardware raid. I have a very simple W2003
> > setup with 5 users on Terminal Services.
> >
> > I want to upgrade my 3YO IBM X206 with a new server that has hardware raid
> > and 2 drives. The new server has cold-swap drives feature. (I cannot get
> > any
> > technical person from IBM to talk to me - go figure)
> >
> > So what happens when a drive fails?
> > Does the W2003 know about? Does it issue a message? Does it stop working
> > altogether until whatever is done?
> > Should I purchase a 3rd IDENTICAL drive so when one fails it can take
> > over?
> > Does it takeover automatically? or do I replace the fallen over one with
> > it?
> >
> > OK you see I have no experience at the recovery end. I have spent a lot of
> > hours scouring the IBM site in order to find out how raid works from
> > beggining to end. No luck.
> >
> > --
> > Regards
> > Tom

>
> You're talking about a disk mirror. When one of your disks fails
> then the system generates an error message but it continues
> operating normally. This gives you the opportunity to shut down
> the system, replace the failed disk and rebuild the mirror. There
> is no automatic failover in a RAID1 configuration.
>
> To gain confidence, you should go through a practice run with
> you mirror. Unplug the data cable and watch what happens (but
> switch the system off before you plug it back in again!). See
> if you can boot the machine with one or the other mirror disks.
>
>
>
 
Re: How does RAID 1 work in restore situation - Briefly.

With "unplug" I mean just this: Pull out the IDE cable. Do it with
a quick, firm move.

While it is correct that the original disk may no longer be available
four years down the track, it is even more likely that you will need
larger disks by then. At that stage you would have several options if
one of your mirrored disks should fail:
a) Replace it with an identical disk, or
b) Replace it with a larger disk, or
c) Replace both disks with larger disks.
d) Continue with a single disk.

Keep in mind that modern disks are very reliable and that a failure
is quite unlikely. File system corruption or a failure of Windows
(perhaps due to some registry problems) are far more likely. You
should therefore ask yourself: What would you do if Windows won't
boot up? Your mirrored disks won't help you here. Do you have
a mechanism in place to recover from this situation, with a minimum
of cost or downtime?

"ThomasAJ" <ThomasAJ@discussions.microsoft.com> wrote in message
news:C9B5CFB7-FA71-4FD8-8B6F-5DBDFC4508E3@microsoft.com...
> When you say "unplug the data cable" can you please elaborate on that.
>
> Also should I purchase an identical 3rd. drive. Bear in mind that failure
> may occur 4 years down the track and the original drive may not be
> available
> and if it is then shipping may take a while during which time there will
> be
> no RAID 1.
>
> --
> Regards
> Tom
>
>
> "Pegasus (MVP)" wrote:
>
>>
>> "ThomasAJ" <ThomasAJ@discussions.microsoft.com> wrote in message
>> news:FBBEBA70-5D7F-4403-BA0A-47DFEC2EE53D@microsoft.com...
>> >I know Raid 1 uses 2 identical disks and stores the data identically on
>> >each
>> > one.
>> > I also only want to talk about hardware raid. I have a very simple
>> > W2003
>> > setup with 5 users on Terminal Services.
>> >
>> > I want to upgrade my 3YO IBM X206 with a new server that has hardware
>> > raid
>> > and 2 drives. The new server has cold-swap drives feature. (I cannot
>> > get
>> > any
>> > technical person from IBM to talk to me - go figure)
>> >
>> > So what happens when a drive fails?
>> > Does the W2003 know about? Does it issue a message? Does it stop
>> > working
>> > altogether until whatever is done?
>> > Should I purchase a 3rd IDENTICAL drive so when one fails it can take
>> > over?
>> > Does it takeover automatically? or do I replace the fallen over one
>> > with
>> > it?
>> >
>> > OK you see I have no experience at the recovery end. I have spent a lot
>> > of
>> > hours scouring the IBM site in order to find out how raid works from
>> > beggining to end. No luck.
>> >
>> > --
>> > Regards
>> > Tom

>>
>> You're talking about a disk mirror. When one of your disks fails
>> then the system generates an error message but it continues
>> operating normally. This gives you the opportunity to shut down
>> the system, replace the failed disk and rebuild the mirror. There
>> is no automatic failover in a RAID1 configuration.
>>
>> To gain confidence, you should go through a practice run with
>> you mirror. Unplug the data cable and watch what happens (but
>> switch the system off before you plug it back in again!). See
>> if you can boot the machine with one or the other mirror disks.
>>
>>
>>
 
Re: How does RAID 1 work in restore situation - Briefly.

Thank you. You said "File system corruption or a failure of Windows
(perhaps due to some registry problems) are far more likely. "
That's a worry as I have been mainly focusing on hardware.

However I have searched for backup solutions and they are usually very
complex as their target is large corporations.
This server is used by 6 people, 1 app over Terminal Services.
Would the BACKUP utility on W2003 be sufficient to store the WHOLE system on
an external USB harddrive which would then be used to boot from?
--
Regards
Tom


"Pegasus (MVP)" wrote:

> With "unplug" I mean just this: Pull out the IDE cable. Do it with
> a quick, firm move.
>
> While it is correct that the original disk may no longer be available
> four years down the track, it is even more likely that you will need
> larger disks by then. At that stage you would have several options if
> one of your mirrored disks should fail:
> a) Replace it with an identical disk, or
> b) Replace it with a larger disk, or
> c) Replace both disks with larger disks.
> d) Continue with a single disk.
>
> Keep in mind that modern disks are very reliable and that a failure
> is quite unlikely. File system corruption or a failure of Windows
> (perhaps due to some registry problems) are far more likely. You
> should therefore ask yourself: What would you do if Windows won't
> boot up? Your mirrored disks won't help you here. Do you have
> a mechanism in place to recover from this situation, with a minimum
> of cost or downtime?
>
> "ThomasAJ" <ThomasAJ@discussions.microsoft.com> wrote in message
> news:C9B5CFB7-FA71-4FD8-8B6F-5DBDFC4508E3@microsoft.com...
> > When you say "unplug the data cable" can you please elaborate on that.
> >
> > Also should I purchase an identical 3rd. drive. Bear in mind that failure
> > may occur 4 years down the track and the original drive may not be
> > available
> > and if it is then shipping may take a while during which time there will
> > be
> > no RAID 1.
> >
> > --
> > Regards
> > Tom
> >
> >
> > "Pegasus (MVP)" wrote:
> >
> >>
> >> "ThomasAJ" <ThomasAJ@discussions.microsoft.com> wrote in message
> >> news:FBBEBA70-5D7F-4403-BA0A-47DFEC2EE53D@microsoft.com...
> >> >I know Raid 1 uses 2 identical disks and stores the data identically on
> >> >each
> >> > one.
> >> > I also only want to talk about hardware raid. I have a very simple
> >> > W2003
> >> > setup with 5 users on Terminal Services.
> >> >
> >> > I want to upgrade my 3YO IBM X206 with a new server that has hardware
> >> > raid
> >> > and 2 drives. The new server has cold-swap drives feature. (I cannot
> >> > get
> >> > any
> >> > technical person from IBM to talk to me - go figure)
> >> >
> >> > So what happens when a drive fails?
> >> > Does the W2003 know about? Does it issue a message? Does it stop
> >> > working
> >> > altogether until whatever is done?
> >> > Should I purchase a 3rd IDENTICAL drive so when one fails it can take
> >> > over?
> >> > Does it takeover automatically? or do I replace the fallen over one
> >> > with
> >> > it?
> >> >
> >> > OK you see I have no experience at the recovery end. I have spent a lot
> >> > of
> >> > hours scouring the IBM site in order to find out how raid works from
> >> > beggining to end. No luck.
> >> >
> >> > --
> >> > Regards
> >> > Tom
> >>
> >> You're talking about a disk mirror. When one of your disks fails
> >> then the system generates an error message but it continues
> >> operating normally. This gives you the opportunity to shut down
> >> the system, replace the failed disk and rebuild the mirror. There
> >> is no automatic failover in a RAID1 configuration.
> >>
> >> To gain confidence, you should go through a practice run with
> >> you mirror. Unplug the data cable and watch what happens (but
> >> switch the system off before you plug it back in again!). See
> >> if you can boot the machine with one or the other mirror disks.
> >>
> >>
> >>

>
>
>
 
Back
Top