So, it happened for us also :(

Due to a series of unlikely events our oldest and largest Exchange Public 
Folder information store got corrupted and went down with a bang. According to 
the on-site Microsoft Premium Field Engineer the store is fubar so we have 
started a restore. The size of the store is 3.5 TB (!), and the restore started 
out pretty well, at the speed that we would expect from the combination of 
server, storage and network.

However, after a few hours restore speed began to slowly decrease, and here 
after almost 24 hours it is down to a crawl. No errors to be seen anywhere, no 
network, CPUs or storage systems are maxed out - the restore just continues 
flawlessly, but at something like a tenth of the expected speed.

TSM client and server software are rather old, Exchange TDP is ver. 5.2.1.0 and 
TSM server is ver. 5.5.2.1. Hardware is 2-3 years old and giga-bit connected. 
We do not see this as an hardware issue.

Anyone seen something like this before and does anyone have any of idea of why 
the restore runs at turtle speed?

- Bent