Archive for November, 2009


during our migration to 2010 from exchange 2007 we had a fairly seamless transition but there were a few little things i would watch out for in the future

1. when moving mailboxes it generates a lot of transaction logs which eat disk space VERY fast so be sure to do small blocks at a time and keep you logs on the separate disk they are supposed to be so it doesn’t kill your database

2. we found that if we rebooted the exchange 2010 servers that the exchange 2007 would build up a queue of mail to it, even when it came back online, a quick restart of the transport service on the cas server sorted it out

3. if you database is say 52gb in 2007 it does not mean it will be 52gb in 2010 directly after the move. so basically clear down everything unessential and if possible defrag the database and then move. why take “junk” mailboxes across.

4. HA in exchange 2010 and 2007 are completely different! i knew about this already but from when i seen prelease builds of 2010 you could do dag with 2 machines and now you need 3! 2 mailbox servers (doesn’t matter if they have the other roles on it) and one independent CAS / HUB Transport server with no mailbox role.

5. The other thing that came up within HA scenarios is can you use Windows Network Network Load Balancing as long as you are not using windows cluster technologies as they cant exist on the same box BUT you can use a hardware load balancer at that stage!

6. Co-existence Works! but you can only manage a 2010 server from a 2010 management console and a 2007 server from a 2007 console. BUT you can view queues of each server using the queue viewer tool. i know this might seem a little bit stupid to say but i have had people in the past asking me about 2003 and 2007 so its best to mention it here!

7. Learn Powershell – 🙂 i know i know if you don’t already at least get familiar with the |fl part of any command for example “get-mailboxserver |fl” the reason i say that is the amount of info |fl will give you is amazing compared to leaving it out.

8. Archiving – fantastic that it is here BUT BIG caveat for now (unless you like beta software like me) it can only been seen in outlook 2010 clients or owa 2010 clients which makes it pretty unusable for now unless as i said you dont mind beta software or the web client which ROCKS!

9. CALS – these have changed in 2010 and are now additive as in you need standard cal for everyone and if you want enterprise features you need an ADDITIONAL or ADDITIVE enterprise cal ASWELL…. don’t be mistaken. one nice feature though in 2010 is a license viewer which will tell you exactly what you need when your deployment is finished….

10. you can DAG with exchange 2010 standard but you need windows enterprise underneath so dont get confused! the main difference between exchange standard and enterprise is the amount of databases you can have, in standard its 6 in enterprise is a big number (i cant remember!) database limit is around 16TB so in most cases standard will probably be more than sufficent


this is all for now.. i think



check out this article its very interesting to backup the virtual machines using CLI Tools

its up on MSDN for those who have subscriptions… mines on the way down 🙂

OCS Resource Centre


OCS update resource centre…


very useful little site



i tell you this has been the week of the most unique problems!


i have seen this issue but cant find any reference on the web, when you change the memory assigned to a windows 2003 sp2 x32 guest to above 4gb the server generates the lsass.exe messages and tries to shutdown the entire virtual machine HOST

once we changed it back to under 4gb it stayed up…. very strange and very much looked like a virus


and before anyone says… yes i know a 32bit host cant see more than 4gb of memory but even then it should crash a host either 🙂


ok, this problem drove me nuts!

basically we have a windows 2008 r2 cluster running live migration connected into a fiber channel eva4400 san

live migration is configured but doesnt work.

i try to access the cluster shared volume on both nodes and it doesnt work on one node with the only message coming back as “Access is Denied”

however a basic active / passive failover cluster works…. confusing…

hp dsm and mpio installed and configured.

host settings set in the san

all the pre-reqs for the nics etc… and configuration of windows for clustering and live migration done

everything in the right order!

so what is the answer…. SYMANTEC ENDPOINT PROTECTION… removed it and up came everything

still working on a work around but so far nothing…. more to post when i have it!