• Normal redundancy (double-mirroring) option providing 6TB of usable storage
• Enhanced Diagnostics - Trace File Analyzer and ODACHK

Also, if you review the README, you may see content that says:   
    "The grid infrastructure and database patching, both are rolling upgradable. During our patching, we patch the node 1 first and when completed, we patch the node 2."

I would like to clarify that the 'infrastructure' updates (OS, Firmware, ILOM, etc) will require a  short downtime of the ODA while it is applied.  When you update the grid infrastructure (--gi), the appliance manager verifies that the infrastructure was updated so you cannot just patch the GI without first updating the infrastructure. The high level update patch steps include (but not limited to):

• Download patch update to your ODA
• The --infra (infrastructure) is updated and ODA Databases are down and the ODA is/may be rebooted
• ODA and GI/Databases are restarted
• Issue the command to update the Grid Infrastructure/databases (The order of the steps are completed automatically and you cannot control when the nodes are brought up and down during the patching)
  
• Node 1 -- shutdown databases and GI
• Node 1 -- patch GI/database
• Node 1 -- bring up databases and GI
• Node 2 -- shutdown databases and GI
• Node 2 -- patch GI/database
• Node 2 -- bring up databases and GI

A replay from Friday's with Sohan on the 2.4 release can be found here.  The PDF of the presentation is here.