2011-06-22 10:31 AM
I have 2 2007 exchange servers on two different domains and want to migrate them to netapp fas2020.
Could you help me with the following (didn't manage to find the appropriate answers in admin guide).
1. Each Excahnge server has around 80 GB of data. What size of LUN and fractional reserve do you recommend?
2. The test install I made asked me to put each public and personal edb and each log on a different LUN (I couldn't
select the same destination for each folder). Did I do something wrong (for 2 servers that would be approx. 8 LUNs)?
3. Should I put each LUN on a different volume?
4. I plan to use SME for everyday backups and use windows ntbackup (copy type) backup to tape. Is this a good policy?
5. Does the previous question affect the first question about sizing of the LUNs and the amount of fractional reserve?
6. Is there any other comprehensive source besides the admin guide that can help me with the planning?
Solved! SEE THE SOLUTION
2011-06-24 11:55 AM
Here's my two cents, may work for you may not... Hopefully it helps regardless.
Point 1 -- If you do information store maintenance on a schedule I would say start your MDB LUN's off at 200GB. The 80GB of data you already have, 20GB to grow into, and 100GB for temp space for maintenance. I recommend creating a log LUN large enough to store 3 or 4 days worth of transaction logs. In the event you run into any issues with backups and aren't truncating these logs for a few days, your LUN won't offline and take mail down until someone notices. With that amount of data I would guestimate 20-25GB would be plenty but you'd know better than I how much mail you move on a day to day basis. Take the combined size of the LUN's you intend to create and multiply by 2.2 for your volume size.
Point 2 -- You should separate database and logs onto different LUN's. SME just enforces adherance to that best practice.
Point 3 -- This one is a combination of personal preference and the load on your system. If the load on the aggregate you put the volume in is low, or it's a brand new aggregate for Exchange, I would say it's fine.
Point 4 -- I know of a few places that have a similar strategy (SME daily or multiple times per day, weekly ntbackup/netbackup to tape for off-site retention). If it meets your needs in the event of a disaster then it's a good policy. You can also syncmirror your data locally for a second copy readily available and/or snapmirror it remotely to handle off-site.
Point 5 -- 20% is a good number for your snap reserve. Whether or not you need to grow that number depends on the rate of change in your environment and how many snapshots you want to keep.
Point 6 -- The admin guide is pretty thorough. I would suggest being thorough as you go through it as it will help you out quite a bit. I would even say re-read it several times just to make sure you don't miss anything. (I do that.)