Hi All, We use NetApp storage for Exchange. Exchange server is connected to NetApp storage via iSCSI. We use the second server's NIC for that. No DAGs, but we have backup. We migrated from Ex 2007 to Ex 2010. I created smaller databases for Exchange 2010 than Exchange 2007. Now there are 7 mailbox databases + PF database, total size is approx 900Gb. Each DB is located on separate lun. I.e. there are created 8 luns for DBs and one lun for logs. Each edb file is approx 120Gb. Ex2007 had 3 mailbox databases and PF db. There were 4 luns for strorage groups + logs lun. Each mailbox DB was approx 300Gb. I compared iSCSI interface usage and see, that Exchange 2010 needs more bandwidth in average than Exchange 2007. I see 1.5-2 times more iSCSI interface usage. My questions: Is there dependency between number of databases and total disks usage ? Will bigger databases in size decrease total disk usage / iSCSI traffic ? I would like to see less interface usage (and storage). Thanks mcse^4

In fact, some of the I/O, the online database management consumes significant bandwidth that wasn't present in Exchange 2007, and that load is proporational to the number of databases because it has a I/O thread for each database. I don't know that this is something you need to be concerned about, however, since you didn't supply any evidence that you're getting anywhere close to I/O issues.Ed Crowley MVP "There are seldom good technological solutions to behavioral problems."

All, can anyone check it too ? mcse^4

Hi Have you verified the I/O with Jetstress? It's always a good practice before deploying into prod When you're running the Jetstress tests you can also check for the network usage as well Jonas Andersson | Microsoft Community Contributor Award 2011 | MCITP: EMA 2007/2010 | Blog: http://www.testlabs.se/blog | Follow me on twitter: jonand82

You will be generating more logs in 2010 than you did in 2007. That will add IO. Ed also told you about the database maintenance. If youre in a migration process or you are working the database a little harder you are going to see added I/O. It will calm down. By all means run the Jetstress but be careful of the results if your NetApp system has a Flash Cache card in it. Your results will potentially look a little odd to someone who might know Exchange inside out but doesnt know NetApp or general network storage. "Jonas Andersson [MCITP]" wrote in message news:8a9277d3-340f-4ec4-8d5b-49e61ecbac24... Hi Have you verified the I/O with Jetstress? It's always a good practice before deploying into prod When you're running the Jetstress tests you can also check for the network usage as well Jonas Andersson | Microsoft Community Contributor Award 2011 | MCITP: EMA 2007/2010 | Blog: http://www.testlabs.se/blog | Follow me on twitter: jonand82 Mark Arnold, Exchange MVP.