During a mailbox move of about 130 people one standalone server to a cluster, about half of the mailboxes (70 give or take) were left behind on the source mailbox server, presumably because of a corrupted item or mailbox. We thought all mailboxes had been moved initially because the dialog box said successful. Further inspection revealed that many of the mailboxes still existed on the source server.
Over the course of this mailbox migration project, we have had several instances where a mailbox is moved, but a small mailbox with several items is left behind (a stub mailbox). By lowering the mailbox retention period and running the cleanup agent, we have been able to get rid of those stub mailboxes.
In this case, we have several stub mailboxes on the destination server, with a portion of the mailbox sitting on the source serve, and don’t know why. In some cases, we have several hundred MB or even a GB on the destination (stub) mailbox. Could it be that something interrupted mailboxes being moved and some of the content exists on the destination server as well as the source server?
We are a little nervous about lowering the retention period and running the cleanup (which is the typical fix action for stub mailboxes) on the destination mailbox server. This problem is the diametric opposite of what we are used to seeing.
The environment is a Windows server 2003 running Microsoft Exchange server Build 6.5.7638.1 on both the source standalone server and the destination cluster server.
I had set the point assignment at 125, as I am not too familiar at gauging point values on here.