If this is 2010-sp1 then step 5 doesn't happen. Logs are sent via TCP write-by-write to the destination DAG server [part of step 3] (until or if that connection is broken, in which case normal log transport takes over).
What you are presenting as a single read/write may be multiple read/writes. If you can only do 5 at a time, then only do 5 at a time. By default you can only do 5 per CAS, but you can control that to some degree. See http://thoughtsofanidlemind.wordpress.com/2010/12/03/tweaking-the-mailbox-replication-service-configuration-file/ Regards, Michael B. Smith Consultant and Exchange MVP http://TheEssentialExchange.com -----Original Message----- From: Stringham, Steven [mailto:[email protected]] Sent: Friday, June 24, 2011 8:13 PM To: MS-Exchange Admin Issues Subject: Exchange 2003/2010 migration - disk access question I have a 2003/2010 mixed environment. I have moved a bunch of mailboxes successfully between the two (both ways). I still have a bunch to move from 2003->2010. What I need to do is optimize the disk writes/reads for the data as it moves. When I move too many mailboxes at time the system keeps getting to the 95% point, and then can't quite seem to complete the move in a timely manner. If I limit the move to under 5 mailboxes at the same time, it goes well... So, the question is, what all disk reads/writes occur on the 2010 side as a mailbox move occurs? My environment: 2003 - multiple servers 2007 - One Front End/One Mailbox (these are there only because I was doing some testing - and users are now using the 2007 OWA as a front end to the 2003 OWA....). 2010 - Two Mailbox/Transport/DAG servers. Two CAS/Transport. One External Edge Transport. The CAS servers seem to be doing the actual moves from 2003, or are at least in charge. Do they actually write the data to disk at any point? The MBX servers get the data, write it to the local queue???, then post it to the log volume, and then to the database. It then reads the changes, and pushes it to the inactive DAGs, who write it to the local queue, logs, and then database, right? So: 1) CAS queue Write/read 2) MBX1 - queue Write/read 3) MBX1 - log Write 4) MBX1 - DB Write/Read 5) MBX2 - queue Write/Read 6) MBX2 - log Write 7) MBX2 - DB Write So, I see 11 Reads/Writes total? Do I have too many? Am I missing any? Can this be optimized in any way? Am I way off base? Thanks Steven Stringham ---------------------- For more information about Lewis and Roca LLP, please go to www.lewisandroca.com. Phoenix (602)262-5311 Minden (775)586-9500 Tucson (520)622-2090 Albuquerque (505)764-5400 Las Vegas (702)949-8200 Silicon Valley (650)391-1380 Reno (775)823-2900 This message is intended only for the use of the individual or entity to which it is addressed. If the reader of this message is not the intended recipient, or the employee or agent responsible for delivering the message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this message is strictly prohibited. If you have received this communication in error, please notify us immediately by replying to the sender of this E-Mail by return E-Mail or by telephone. In accordance with Internal Revenue Service Circular 230, we advise you that if this email contains any tax advice, such tax advice was not intended or written to be used, and it cannot be used, by any taxpayer for the purpose of avoiding penalties that may be imposed on the taxpayer --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to [email protected] with the body: unsubscribe exchangelist --- To manage subscriptions click here: http://lyris.sunbelt-software.com/read/my_forums/ or send an email to [email protected] with the body: unsubscribe exchangelist
