Solved

Exchange 2007 CCR cluster - best way to swap storage on the passive node

Posted on 2010-11-08
18
776 Views
Last Modified: 2012-05-10
I have approx. 15 storage groups in a CCR cluster. I used local storage for the passive node when it was first implemented, but now we have SAN storage. What's the best way for me to move databases and logs on the passive node from local storage to SAN storage? I'm thinking that since CCR takes care of replication I could simply do the following; 1) stop replication between the nodes, 2) swap the drive letter between the local storage and the SAN storage, 3) resume replication. I think this will copy all 15 database files over along with the appropriate logs (it will take sometime since each database is approx. 80GB and I have 15 storage groups). My other idea was to do #1 and #2 but instead of letting the replication copy the database files I would copy the files over manually. Not sure if this provides an advantage at all but it was just another idea. Anyway, appreciate it if you can give me some feedback. Thanks.
0
Comment
Question by:bndit
  • 9
  • 6
  • 3
18 Comments
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34090931
I reckon a Reseed is your best option, but you could try the  suspend, copy files and resume, but I think letting Exchange Reseed the copies would be quicker.

Have you considered Mount Points?
0
 
LVL 49

Expert Comment

by:Akhater
ID: 34090938
oh well you can

1) suspend replication
2) swap drive letters
3) copy paste from the local disk to the new drives
4) resume replication

this will bypass the initial seeding time
0
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34090970
@Akhater - I think the copy/paste would take as long as reseeding and is more prone to human error...
0
 
LVL 49

Expert Comment

by:Akhater
ID: 34090975
@MegaNuk3 good point at second thought I agree with you.

if both servers are on the same back bone / fast network then reseeding would be better

0
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34091026
@Akhater - thanks, it is nice to know my brain works Pre caffeine...
0
 
LVL 2

Author Comment

by:bndit
ID: 34096606
Ok, so I think reseeding the databases is the way to go. Now, my options are 1) Automatically, 2) Update-StorageGroupCopy cmlet, and 3) Manually. I think option #2 applies in my case as it cannot be automatic anymore and manually copying might be a good idea. My question now...since I have 15 SGs and the location of the database and log files must be the same between the cluster nodes, this means that I have to reseed all 15 SGs at the same time, correct? Here's the MS KB article I'm looking at: http://technet.microsoft.com/en-us/library/bb124706%28EXCHG.80%29.aspx
0
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34097553
You can suspend them all then resume them one at a time to let them seed and synch up 1 by 1
0
 
LVL 2

Author Comment

by:bndit
ID: 34097709
@MegaNuk3 - I'd love to do that but my point is that I can't do that since I'm switching the drive. Let me explain further...

Currently Node2 has a local storage array with letter D:\. That's where all my 15 SG's go for database and log location. Since I'm swapping letter drives with the new SAN storage..this new storage will have D:\ and the local storage will have some other letter drive, say N:\. Node1 will no longer see any of the 15 SGs anymore so that's why I say that I'm forced to do them all at the same time. If I'm misunderstanding you or if there's anything that I can do so that I can throttle the number of SG's please share with me.
0
 
LVL 31

Accepted Solution

by:
MegaNuk3 earned 175 total points
ID: 34097825
Suspend the SG copy for all 15 then swap the drive letter on the passive node, then resume/update the SGs one by one. So you watch them go from Seeding to Healthy status.

If the copy is suspended then the active node won't care what is happening on the passive node until the copy is resumed/updated.
0
 
LVL 2

Author Comment

by:bndit
ID: 34097885
Ok, I'm with you so far up until the resuming part. Are you suggesting that I resume one SG at the time...does that mean one after the other? or, resume for one...wait until it's completely reseeded and then start another one? If the latter, this will require a lot of babysitting and time on my part...if that's the case I'd do them all at the same time.
0
 
LVL 49

Assisted Solution

by:Akhater
Akhater earned 175 total points
ID: 34099666
it doesn't really matter if one at a time or not, it depends on your network speed and the size of your databases.

resuming 15 SG together will mean a lot of network traffic it might lead to a bottleneck however you can resume 2 or 3 together wait till they finish and then enable other 2 or 3
0
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34099742
As Akhater says, you are restricted by the network speed.

You could always ramp it up, do 1 so you know the process, then do 3, see how long that takes, then do 5 then do the final 6.

If it was me I'd do 1 by 1 to make sure each SG reaches a Healthy status before moving onto the next one. Doing one by one will probably take the same amount of time as doing all 15 at once
0
 
LVL 2

Author Comment

by:bndit
ID: 34105609
Ok fair enough. I read that reseeding could impact client access so since I'm doing 15 SGs @ 80GB each, I'll go ahead and start it on Friday evening and continue through the weekend. I'll go ahead and close this post over the weekend. Thanks for all the help.
0
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34106047
Good luck, hope it all works for you
0
 
LVL 2

Author Closing Comment

by:bndit
ID: 34131974
Thanks guys. You both were helpful. It worked like a charm.
0
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34132030
Thanks for the points. Did reseeding all 15 take long?
0
 
LVL 2

Author Comment

by:bndit
ID: 34132045
No problem, thank you for responding to my question. Each mailbox database took a little over an hour. The SAN is on a different LAN but has a 1Gbps connection to the CCR node. I started on Friday evening and finished Saturday afternoon. I tried to automate the process but it didn't work out too well for me so I decided not to spend more time tweaking the script that I had....totally doable though so maybe I'll work on it in a test environment just to have it..in case I need to do the same thing.
0
 
LVL 31

Expert Comment

by:MegaNuk3
ID: 34134158
Hopefully you won't have to do the same thing again, but at least you know how to do it.
0

Join & Write a Comment

Possible fixes for Windows 7 and Windows Server 2008 updating problem. Solutions mentioned are from Microsoft themselves. I started a case with them from our Microsoft Silver Partner option to open a case and get direct support from Microsoft. If s…
Scam emails are a huge burden for many businesses. Spotting one is not always easy. Follow our tips to identify if an email you receive is a scam.
This tutorial will walk an individual through the steps necessary to enable the VMware\Hyper-V licensed feature of Backup Exec 2012. In addition, how to add a VMware server and configure a backup job. The first step is to acquire the necessary licen…
To add imagery to an HTML email signature, you have two options available to you. You can either add a logo/image by embedding it directly into the signature or hosting it externally and linking to it. The vast majority of email clients display l…

759 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question

Need Help in Real-Time?

Connect with top rated Experts

21 Experts available now in Live!

Get 1:1 Help Now