Hopefully somebody is still monitoring this thread; I have some information that you might like to hear. My name is Christian Allred and I'm a developer at Microsoft on Transactional NTFS (TxF). TxF is the component responsible for the transactional resource managers mentioned in the NTFS 137 events. I just wanted to let you guys know that we've fixed this issue for Vista service pack 1 and Windows Server 2008. Acronis provided us a copy of TrueImage, and we were able to determine that the issue is a bug in an NTFS routine that handles the FSCTL_LOCK_VOLUME control code. The bug was in the mechanism that we use to get TxF resource managers out of the way to allow a LOCK_VOLUME request to work. The result of a bug was that if sombody used an asynchronous handle to issue FSCTL_LOCK_VOLUME, we'd erroneously keep a flag set that would prevent us from starting TxF resource managers again on the volume until you dismounted/remounted the volume. It turns out that some backup and defrag products, including TrueImage, issue FSCTL_LOCK_VOLUME on async handles, triggering the bug. The reason that putting a page file on a volume prevents the bug (and cause the "Analyzing partitions" part to move faster, as mwang noted) is that TrueImage doesn't bother to try locking the volume, since it knows already that the lock will fail in the presence of a page file. The info you guys provided in this thread was quite useful to us in narrowing down the cause and figuring out how to fix this. Thanks.