Planning the Storage Server Disk Group Layout

In general, follow this process to determine how to lay out your zones and clusters across the storage servers, where each decision point at the top of the process affects decisions made lower in the decision tree.

1. Determine the amount of existing usable disk space. How many storage servers are available for the clusters that you are creating? How much total usable disk space is available for those clusters?

   For more information, see Determine the Available Existing Storage Capacity.

2. Determine the number of clusters that you want for the zones that you are creating on the Database Domains. This, in conjunction with the total amount of existing usable disk space, is used to determine the maximum amount of space that will be available for each cluster.

   For more information, see Determine the Number of Clusters in the System and Maximum Amount of Space Available to Each Cluster.

3. Determine the amount of space to assign to each of the three disk groups in each cluster.

   • DATA

   • RECO

   • DBFS

   ---

   Note - There is an option to designate the DBFS disk group set as UNUSED, if necessary.

   ---

   Also determine the names to use for the disk groups for each cluster. For example, if you are creating disk groups for two clusters (osc01r1 and osc01r2), then you might use these names for the three disk groups for each cluster:

   • First cluster (osc01r1):

     - DATA_osc01r1

     - RECO_osc01r1

     - DBFS_osc01r1

   • Second cluster (osc01r2):

     - DATA_osc01r2

     - RECO_osc01r2

     - DBFS_osc01r2

Record your storage server disk group layout in the worksheet in Recording Your Storage Server Disk Group Layout.

For more information, see Determine the Amount of Space for Each Disk Group in Each Cluster.

Determine the Available Existing Storage Capacity

You must first determine how much existing storage capacity is available before deciding how to carve that storage up for the disk groups for each cluster that you create.

Every storage server contains 12 drives, with the following capacities:

• 1.2 TB High Performance SAS disk

• 4 TB High Capacity SAS disk

In addition, the total number of storage servers varies, depending on the type of Exadata Storage Expansion Rack that you have:

• Storage rack (part of the Oracle SuperCluster M6-32): 9 storage servers

• Optional expansion rack:

  • Quarter rack: 4 storage servers

  • Half rack: 9 storage servers

  • Full rack: 18 storage servers

Finally, the total capacity that you have available depends on the level of redundancy that you want to set for the ASM disk groups (normal redundancy or high redundancy).

The following tables provide capacity information for the Oracle SuperCluster M6-32 storage rack and the expansion rack.

Table 37 Storage Server Storage Capacity in the Oracle SuperCluster M6-32 Storage Rack

Table 38 Storage Server Storage Capacity in the Expansion Rack

Determine the Number of Clusters in the System and Maximum Amount of Space Available to Each Cluster

The number of clusters in your system varies, depending on several factors, such as:

• The logical divisions for the databases in your system, where clusters that are created, and zones that are members of those clusters, might have different functions or audiences.

• The number of zones that are available and the number of clusters that can be created with those zones as members, which is partly determined by several factors, such as the number of Database Domains, the size of each Database Domain, and so on.

See Zones and Cluster Planning for Database Domains for more information on how to decide on the total number of zones and clusters in your system.

Once you determine the number of clusters in your system, you can now determine the maximum amount of space that you can assign to each cluster based on the total amount of usable storage that is available. The total amount of space that you assign to all the clusters in the system cannot exceed the total amount of storage that is available.

For example, assume you have a total amount of 48 TB available as usable storage, and you want to create four clusters total in your system, where each cluster has the same overall size. Then the maximum amount of usable space for each cluster would be 12 TB apiece.

Determine the Amount of Space for Each Disk Group in Each Cluster

Once you have decided on the total amount of disk space to use for each cluster, you must break that total disk space for each cluster down into disk space that is used for the three disk groups within each cluster:

• DATA

• RECO

• DBFS

The DBFS disk group is normally assigned a substantially smaller amount of total disk space than the other two disk groups. You can also designate the DBFS disk group as UNUSED, if necessary.

For example, if 12 TB of space is available for each cluster, you must carve that 12 TB of space up so that the combined disk space used by all three disk groups is 12 TB or less.