Caution - Take extreme care when specifying the DATA and RECO disk group names. The names must be unique for the cluster you are defining. Check existing grid disks on the storage cells and ensure that they are not already in use. If an existing grid disk is specified for a new cluster, the griddisk is dropped and recreated by older versions of Java OneCommand (prior to patch 19766190), possibly resulting in loss of production data. |
The left pane displays a cluster configuration page for each new cluster that you are creating.
Grid Infrastructure Home
Database Home Location
In the Diskgroup Layout, chose one of these options:
Legacy 80%:20% – Assigns the following sizes to the disk groups:
DBFS Diskgroup: Default size (the default size for the DBFS disk group in this selection is the size of the operating system disk slice on disks 0 and 1, which is usually between 29 GB and 31 GB)
DATA Diskgroup: 80% size
RECO Diskgroup: 20% size
Caution - Choose Legacy only if you have dedicated storage cells per cluster and no storage cell sharing. |
Legacy 40%:60% – Assigns the following sizes to the disk groups:
DBFS Diskgroup: Default size (the default size for the DBFS disk group in this selection is the size of the OS disk slice on disks 0 and 1, which is usually between 29 GB and 31 GB.)
DATA Diskgroup: 40% size
RECO Diskgroup: 60% size
Caution - Choose Legacy only if you have dedicated storage cells per cluster and no storage cell sharing. |
Custom – Allows you to assign your own sizes to the disk groups
Specify a unique name for the disk groups for this cluster. For this example, these names are specified.
DBFS Diskgroup: DBFS_DG
DATA Diskgroup: DATAC1
RECO Diskgroup: RECOC1
Select the level of redundancy for the disk groups (DBFS, DATA, and RECO):
UNUSED (DBFS disk group only)
NORMAL
HIGH
Configure the size of the disk groups (DBFS, DATA, and RECO) for this cluster.
Management Hostname
10Gb Client Hostname 1
Exadata IB Hostname
VIP Hostname
Hostname and VLAN tag for SCAN Address Group
Admin Network:
Name mask
Client Network:
Name mask
VIP Name mask
SCAN Name
VLAN ID
Private Network:
Name mask
This figure shows how the entries from the two database zones in the Zone Details screens in the SuperCluster Virtual Assistant map to the entries in the networks area at the bottom of the Cluster 1 page in OEDA.
Note the following:
Certain entries in the Cluster 1 page in OEDA have a single field that would be populated using multiple entries from the individual Zone Details screens in the assistant. For example, there is a single Admin Network: Name mask field in OEDA that is populated using the Management Hostname entries from the Zone Details screens in the assistant for each database zone that is part of the cluster, such as etc14m7iozadm010101 for the first database zone and etc14m7iozadm020101 for the second database zone in the assistant. OEDA will use the percentage sign (%) as a wildcard in these cases (in this example, using etc14m7iozadm0%0101 in the Admin Network Name mask field in OEDA).
The following suffixes should be appended at the end of the following entries in the Cluster 1 page in OEDA, even if these suffixes don't appear in the Zone Details pages in the assistant:
-client at the end of the entry used in the Client Network: Name mask field.
-vip at the end of the entry used in the Client Network: VIP Name mask field.
This table shows example entries that might be found in the individual Zone Details pages in the assistant for each database zone and the entries that should be used in the Cluster 1 page in OEDA.
|
Go to Complete the Cluster Review and Edit SCAN, Client, VIP, and Optional Backup Networks Page.