An interesting thing happened while moving mailboxes…

I had to move a few mailboxes today from our old Microsoft Exchange 2003 server to our new 2010 server. When I went to refresh the view in the Move Request section of the navigation pane I got an error

Couldn’t load the request. Not enough information was provided to read the orphaned request message. It was running the command ‘Get-MoveRequestStatistics -Identity ‘DOMAINNAME/OU/Username.

This error threw me for a few seconds. Having noted that sometimes a move request initiated by another admin wouldn’t show up in my console I decided to check out the EMS. I ran the get-moverequest cmdlet and it didn’t list the username the error did.

This left me confused for a a little while. I started checking a couple of things for other issues I’ve seen, including going into ADSIEdit and checking the HomeMTA and HomeMDB values. While in there I noticed something funny about the username that threw the error : unlike another user on the same datastore this user had a non-null value for msExchMailboxMoveStatus. I changed that to <not set> as it was on my working user, and after a few minutes the error went away.

Copyright info

High Availability and Disaster Recovery : They ARE not the same

I have run into people over the years who don’t understand that there is a difference between the two.

I’m going to use a more common technology to explain the difference – run-flat tires. For those not familiar with these they are tires which, even with a puncture of upwards of an inch, can be used for up to 50 miles without being changed out.

Let’s say we stick a set of these on a car. The car now has what could be called ‘high availability’ tires. Run over a nail? No problem – you can keep going for a while without having to worry about losing control of the car. The run-flat tires provide the car with high availability – being able to continue to use the car while in a degraded condition. In the computer world this would translate to various technologies – RAID configurations, clustered servers, replicated configurations, etc. Again, the goal is to provide a level of functioning while in a degraded condition.

How is this different from disaster recovery?

Still using our run-flat tires we’re going to expand the discussion. Say, instead of a nail you run over a  shard from a 2×4 that puts a five inch gash in the tire. The high availability of the run-flat tires is now useless. Now you have to have a spare tire. This is the difference between high availability and disaster recovery : disaster recovery allows you get going again after a total failure. In our tire example the spare tire is what in the computer world is called backup.

Translating this pair of examples into computer world jargon – the high availability is usually best known in clustering and RAID, disaster recovery in backups. HA can protect against system outages, but if Joe Disgruntled decides to delete a bunch of things HA will do no good while DR via backups will.

Copyright info