Introduction:
Exploring the add-on of ACFS was a pleasure. As a pilot it has been implemented now and it looks we got ourselves a new standard which is awesome. With the slides below i would like to offer a quick demo and some suggestions for a standard.
When starting the concept of ACFS it was first thought to set it up as one huge Disk group and all Databases both being Racs and Non- Racs to write into them. In order to lower risks (never know who tries to steal an available disk on the admin level ( been there , have suffered from that i mean)) it has been settled to set up the ACFS on a per database basis in a separate Disk Group in ASM.
Second step in minimizing the Risks is that even though we are talking about ACFS only being a mount point for log files ( diag_destination ) and for the logs of the local listener ( also dedicated on a per database basis ) we have set it up in Normal Redundancy. ( Smiles given the cost of storage these days that only sounded more than fair).
ACFS has been implemented on Linux Red Hat EE 5 , with Grid Infra 11.20.3 and Rdbms 11.2.0.3. in ASM environment.
Note with ACFS as a new standard , we can move toward the Rac Ones if needed more easily (some customers want only a single Instance/DB ). The use of that technology and the hows and why , well that will be a future post.
Anyhow below is the power point.
As always happy reading.
Mathijs
Mathijs Bruggink 