Re: Backup question from MCSE
- From: "Mark Arnold [MVP]" <mark@xxxxxxxx>
- Date: Sun, 19 Nov 2006 18:42:46 +0000
On 16 Nov 2006 05:20:52 -0800, "pandi" <k.pandi.k@xxxxxxxxx> wrote:
Hello folks, I had recently appeared for Exchange server exam...I don t
know answer for this question....I will be happy if you answer this
question
The ServiceLevelAgreement for a company requires Failed Exchange Mail
store must be online with in 1 hour or less. It also says all email
data should be retained for atleast one year.
So, there are a couple of ways to do this; either use the RSG to
restore and repair the store and mount an empty database to get the
users up and running, then backfill the store with repaired/recoverred
data. You can also use a SAN such as a NetApp appliance to take
regular snapshot backups and restore them.
The 2nd option has its own problems because you need to run the
eseutil on the snapshot (either at the time of taking it or the time
of restore) which isn't so bad because the SAN will do that under the
covers as part of the VSS.
The company uses 2 mail box store named MBX1 and MBX2. Both mailboxes
reside in Storage Area Network. MBX 1 is 25 GB. MBX 2 is 22 GB. There
is 153 GB available disk space in the storage area network. You back up
and restore the company mailat a rate of 12 GB per hour. You need to
ensure you should meet the SLA?
If you have issues with your restore times you can add another 18 more
stores. Memory requirements will be greater but backup times will be
quicker and more importantly so will restore times. You can restore
and recover 4 stores at once (in different SG's as store restores in
an SG are serial)
What should you do ?
A) Every night perform full backups to tapes and archive it. Then
perform a shadow copy backup to the storage area network.
"To" the SAN ? "On" the SAN surely? Did you mean to say that or are we
storing Exchange on DAS
B) Perform full backups to tapes on Saturday night and archive it.
Perform differential backup to tapes from every Sunday through Friday.
C) Perform full backups to tapes on Saturday night and archive it.
Perform incremental backups to tapes from every Sunday through Friday
night.
That won't make much difference to the restore times really. Do a full
backup each day if you can and if you're on DAS so that your tape
doesn't spend time messing around seeking. If you're in the position
you think you are you'll have a challenge around even the minute or so
it takes to find the incremental/diff logs.
D) Enable circular logging for the storage groups. Every night perform
shadow copy backups to the storage area network and copy the shadow
backups to tapes. Archive the tapes.
(D) ohhhhhhh no, not even in jest. First off, if you run CL on a
mailbox server I shall come over and pour weak American beer on you (I
won't waste good English beer) and secondly it won't make a difference
to the backup or restore speeds.
.
- References:
- Backup question from MCSE
- From: pandi
- Backup question from MCSE
- Prev by Date: Re: Query-Based Distribution Groups
- Next by Date: Re: Query-Based Distribution Groups
- Previous by thread: Backup question from MCSE
- Next by thread: Two Exchnage Servers on same network
- Index(es):
Relevant Pages
|
