Drobo: all 25 reasons (so far) not to put one into production

Yesterday I posted “Reason 25 not to use a Drobo” and refereed to an earlier post listing another reason. Some folks have asked what the other 23 are. My running list so far, in no particular order: (note this is specific to the DroboElite, I haven’t had the misfortune of playing with the rest of their business line yet)

no redundant power supply
no redundant controller
no redundant NIC (two build-in NICs share a controller)
updating firmware requires loss of access
no one point of contact for support (drive manufacturer/Drobo)
no snapshots
no replication
no resizing volumes
no restricting access by IP address/range
no restricting access by iqn
no VLAN
no LAG/LACP/trunking
only one local user
no performance statistics
no troubleshooting capabilities
no user-readable log files
no OOB management (you configure with the USB then it never works again?)
very long rebuild times
very limited management
changing formats on a volume toasts the unit
no SNMP
no phone home
one email alert (can’t send “all” to one user and “important” to another)
expensive for what you get
And Reason 25:the accepted way to change formats is to copy all the data off