high availability and DR architecture

[Roshan]

Hi,

We have a production environment (2 nodes RAC with grid ASM 12c), OS: Red Hat Linux.

We are planning to implement a DR(Dataguard) with it.

First option is we use a shard VSAN storage
DR.png

Second option is we use extended RAC cluster but this technology does not support full DR.
extendedracversion11-435972.pdf
On page 16, it says look for alternatives to mirroring the storage in addition to a full DR solution.
Should I user VSAN storage for production site and SAN storage for second site(DR) for mirroring?

To implement full DR, the doc says to use RAC One node. How is it different from Extend RAC cluster?

Please advise.

Regards,
ROshan

[ErmanArslansOracleBlog]

Dataguard is the best for Oracle DR.
You can also use storage snapshot/snapmirror -- replication , as an alternative to dataguard, but dataguard is better.  (you have to know what you are doing, and you need to put the database in hot backup mode, before your replication, before your storage snapshot(or any similar thing) is taken)

RAC Extended Clusters can be thought like a disaster solution in a manner, as well. However, it does not protect from human errors or corruptions because the nodes are using the shared storage.

That document is specifically written for Oracle RAC one node.
However, you can built an Extended Cluster for multinode Oracle RAC (not only for Oracle RAC one node), as well.
You can think Oracle RAC one node as a active-passive cluster solution.

Read my blog posts, I have written lots of Dataguard, Storage replication and Extended Cluster related whitepapers.