Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

Long story short: My standalone Exchange 2007 server existed in a Hyper-V instance that somehow clashed with the physical server's RAID hardware and caused a BSOD on the host shortly after the instance had booted up.

I "fixed" the problem by moving the HDDs from the Adaptec HostRAID to be connected directly to the "normal" SATA controller (it was RAID-1, and I wasn't expecting it to be this easy) and then creating a new Hyper-V instance and re-installing WS2008 and then Exchange 2007 with the setup /m:RecoverServer option.

After that finished, I mounted the old instance's VHD and copied the 8GB of ESE Transaction Logs and EDB database files over to the new instance, and told the now running Exchange server that the Mailbox Database files could be replaced by a restore, I then overwrote the "new" database with the old one, then told Exchange to mount it.

...no complaints were displayed in the MMC console and I could access my mail from within OWA (as the server is also a client access server).

This all seems suspiciously simple. Is there anything I might have forgotten or overlooked? My solution was a hack as I hadn't yet implemented any form of backup (yeah, there's about 6000 transaction log files in the database directory) and this was working fine using the simpliest of recovery techniques, yet I understand it's quite painful to recover from an Exchange backup?

share|improve this question
5  
Uncork the champaign! –  Jim Zajkowski Nov 15 '09 at 4:45
    
Just did exactly what you did about 2 weeks ago, and 0 failures. Yes it's that simple as long as the Active Directory is in tact. –  Stephen Thompson Nov 15 '09 at 10:49
1  
It is that simple. My first production Exchange 2007 deployment suffered a catastrophic failure of the RAID-10 array that the database was stored on within a week of deployment (thanks Dell PERC6 somethingorother...). We got a new RAID-10 array put together, restored the last database backup, started the information store, and watched the transaction logs play forward to the point of failure. The database engine behind Exchange (the ESE) rocks, so long as you configure your storage layout properly to begin with. –  Evan Anderson Nov 15 '09 at 13:50
add comment

1 Answer

If no warnings are being thrown, there is nothing else to do than to do a production run. Test run it for a week, do daily backups, and make sure everything is running smooth during this period (ask your users to report any quirks and check logs regularly).

If nothing comes up during that 7 days period, crack open a beer, sit back, and relax!

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.