I've recently had to migrate away from my Linux administration duties to help work on an OS X server install. I've already run through most of the gotchas and had to reformat our server to case insensitive (case ignorant) filesystems all around as well as install the newest version of Server 2.1 directly rather than upgrading. I've been attempting to have a stable OS X server for around 3 weeks on a new mac pro with 4 3TB drives in it and 12 gigs of ram. I have a Primary partition that is a mirrored raid at the start of drives 1 and 3. I have a Secondary partition that is a mirrored raid at the start of drives 2 and 4. I have some 64G partitions on each drive as the second partition for misc stuff.. I store Mountain Lions direct installer on one of the drives so that reinstalling is quick.. I've been doing it a lot. I have a 2TB partition at the end of each drive that is part of a Raid10. I've noticed that this is a literal Raid10 and not a direct Raid10 mapper implementation like I'm used to (where you have near, far, or offset block layouts as an option). All are using HFS+ Journaled. I learned my lesson with Mac-ZFS and Zevo ZFS the first few weeks. Now that I'm doing everything the "Apple" way I expected the system to be rather happy and smug with itself as most Apple things are.. however I had the system die 2 days after a new installation while using ditto to transfer data from an external drive to the newly formatted raid10. Previously I had used a lot of hard links on the files on the external drive. I'm curious if this somehow caused a problem with HFS+ or Ditto. I feel as though it has to have been with HFS+ because the entire machine became unusable. Terminal was no longer usable. 'Top' which was running in one no longer operated. I could move windows around in the U.I. but they wouldn't update.. and of course rainbow wheel of doom just sat there spinning at me the whole time. It may be worth noting that the external drive is case sensitive (case smart), however I scanned it and show no files that would have caused a problem. I'm used to having tools to help me diagnose problems like this. I'm open to suggestions on how to recover from kernel level faults like this on OS X 10.8.x and any ideas on what the potential problem was.