Database Reference
In-Depth Information
At this point, we have gone through the process of creating a vmtemplate and then cloning the vmtemplate to
create a working VM called ol6test. We have assigned High Availability options as well as overridden some values that
we didn't want to take from the template. Now we have a running VM on which we can deploy any application,
as well as an ODA_BASE that is running the Oracle Database. This concludes the most basic configuration required
for running a virtualized ODA.
Advanced VM Configuration and Patching
Configuring visualization on the ODA can be as simple as reimaging the ODA, creating a repository, importing a
vmtemplate, and then cloning it to create a VM. There are advanced configuration techniques, however, that can be
used to provide better control over the environment. We will briefly discuss some of the techniques used to provide
stability and availability to the VM.
CPU Pools
One of the key benefits of the virtualization is controlling various resources and isolating workloads. The default
configuration of creating a VM uses the unpinned CPU pool. This allows access to all the CPUs in the virtualized
appliance to all the VMs , with the exception of ODA_BASE.
ODA_BASE is the Database VM, and as such, when it is created, a pool called
odaBaseCpuPool
is created.
This pool contains the CPUs that only ODA_BASE has exclusive access to. If ODA_BASE exists, you will
have two pools:
•
default-unpinned-pool
•
odaBaseCpuPool
The
default-unpinned-pool
is the pool that contains all the CPUs that are not part of the odaBaseCpuPool. In a
simple configuration, all VMs will have access to the default pool. In cases where isolation is required, you can always
create additional CPU pools to cage a VM. Listing 10-19 shows a configuration of CPU pools.
Listing 10-19.
CPU Pools
#oakcli show cpupool -node 1
Pool Cpu List
default-unpinned-pool [12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23]
odaBaseCpuPool [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,11]
The CPU pools do not have to be uniform across the two nodes—with the exception of odaBaseCpuPool, which
has to be the same across both nodes. Listing 10-20 shows how to create a CPU pool. Listings 10-21 and 10-22 show
how to add CPUs and pin a VM to a CPU pool.
Listing 10-20.
Creating and Managing a cpupool
#oakcli create cpupool unxpool -numcpu 6 -node 0
Listing 10-21.
Add CPUs to a cpupool
#oakcli configure cpupool unxpool -numcpu 10 -node 0
