Just trying out Dave Winers OPML blog/outlining tool (available free for Windows & Mac) - my test site is over here.

More on the joys of ZFS. It actually sounds really really good. I wish I had a pile of hardware to try this stuff out on...

A two part piece on Exchange 12's new features - part one and part two.

Via Adminfoo - Nine principles of Information Security.

Useful - Exchange DR. It all looks horribly painful.

In fact Windows DR in general is horribly painful - if you have a DR setup be sure the hardware is *exactly* the same as whatever you are recovering or a System State / Full-restore (ditto a disk image) will render your machine unbootable (and you need to boot from a Windows CD and try and do a repair). Its interesting to note that many Backup vendors sell Bare-metal restore options to their backup software so the implication is of course that its a PITA to use your backups to restore anything to a non-identical system - which is probably what you'd be doing in a DR situation - as opposed to just doing a server repair / restore on existing server. At least on a Unix system you just restore the config files (mostly in /etc) and the binaries for any applications (usually self contained in their own directory structure) and you're away. Sigh.

The only decent option for failover or DR in Windows would appear to be running a parallel setup offsite with a 'fat-pipe' you can do a nightly synch to (or more regularly for domain controllers / mail etc). I guess if things went really pear shaped its a question of priorities - login, print, file access would be a primary concern but depending on the nature and spread of any failure / disaster it maybe that even getting power will be a problem (or even staff to do do the DR).