Recovery is also pretty useful - I posted about our recovery from a worst-case-total-server-death here: http://unfeatureddocuments.com/content/disaster-and-rec

Recovery is also pretty useful - I posted about our recovery from a worst-case-total-server-death here: http://unfeatureddocuments.com/content/disaster-and-recovery http://unfeatureddocuments.com/content/disaster-and-recovery For a small startup without a dedicated sysadmin I can't think of anything better than EC2 for handling this kind of thing.