The hypervisor present on each Oracle VM Server is an extremely small-footprint virtual machine manager and scheduler. It is designed so that it is the only fully privileged entity in the system. It controls only the most basic resources of the system, including CPU and memory usage, privilege checks, and hardware interrupts.

The terms "domain", "guest" and "virtual machine" are often used interchangeably, but they have subtle differences. A domain is a configurable set of resources, including memory, virtual CPUs, network devices and disk devices, in which virtual machines run. A domain is granted virtual resources and can be started, stopped and restarted independently of other domains or the host server itself. A guest is a virtualized operating system running within a domain. A guest operating system may be paravirtualized or hardware virtualized. Multiple guests can run on the same Oracle VM Server. A virtual machine is a guest operating system and its associated application software. See Section 7.2, “Virtualization Modes (Domain Types)” for information on virtualization modes.

Most of the responsibility of hardware detection in an Oracle VM Server environment is passed to the management domain, referred to as domain zero (or dom0). On x86-based servers, the dom0 kernel is actually a small-footprint Linux kernel with support for a broad array of devices, file systems, and software RAID and volume management. In Oracle VM Server, the dom0 is tasked with providing access to much of the system hardware, creating, destroying and controlling guest operating systems, and presenting those guests with a set of common virtual hardware.

On SPARC-based servers, the management domain, usually referred to as the primary domain, is created when the logical domains manager is installed. If installed on an existing server that is not already configured for logical domains, the current Operating System automatically gets promoted to primary domain status. The primary domain runs an Oracle Solaris kernel and is responsible for the creation and management of all other domains. It is also responsible for providing access to virtualized hardware resources.

On systems running Oracle VM Server for SPARC, aside from the concepts of a "management domain" and of "guest domains", similar to the "user domains" described in Section 2.4.4, “Domains (domU)”, there are a variety of other domain-types that can run alongside the management domain. For instance, it is possible to set up "service domains" that can act as network switches and virtual disk servers. Not all of these domain types are configurable using the Oracle VM Manager.

Guest operating systems each have their own management domain called a "user domain", abbreviated to "domU". These domains are unprivileged domains with no direct access to the hardware or device drivers. Each domU is started alongside dom0 running on Oracle VM Server.

A storage repository is a central location where various resources to build virtual machines are stored. These resources include templates, ISO files, VM files and so on. Oracle VM Servers have shared access to storage repositories for optimized usage of available disk space in the environment, as well as easy reallocation of virtual machines in case a physical server should malfunction.

However, storage in Oracle VM is more than repositories: it also encompasses server pool file systems for clustered server pools, physical disks, or LUNs, in storage arrays, and local physical disks on the Oracle VM Servers. All these storage elements are used in various ways and managed centrally through Oracle VM Manager.

A server pool is a required entity in Oracle VM, even if it contains a single Oracle VM Server. In practice, several Oracle VM Servers will form a server pool, and an Oracle VM environment may contain one or several server pools. Server pools are typically clustered, although an unclustered server pool is also possible.

Server pools have shared access to storage repositories and exchange and store vital cluster information in the server pool file system. In a server pool, a Master server is elected, which is responsible for centralized communication with the Oracle VM Manager. If necessary, any other member of the server pool can take over the Master role. Even in case of server failure, the server pool remains accessible at its virtual IP address, which all servers in the pool share.

Within a clustered server pool, virtual machines can be live-migrated for load balancing purposes or for scheduled maintenance. If a pool member disappears for whatever reason, its virtual machines can be recovered and brought back up on another Oracle VM Server because all necessary resources are available on shared storage.

The networking infrastructure in the Oracle VM environment comprises connections between Oracle VM Servers, between Oracle VM Servers and Oracle VM Manager, between the Oracle VM Servers and their storage sub-systems, as well as communications among virtual machines deployed in the environment, and between virtual machines and external private or public networks. These networking connections can leverage features supported by Oracle VM, such as networked file systems, clustering, redundancy and load balancing, bridging, and support for Virtual LANs (VLANs).

The physical network is the collection of physical connections in Oracle VM Manager and all Oracle VM Servers, and the switches and routers that allow information to reach its destination. A logical network in Oracle VM is built on top of these physical connections. When you create an Oracle VM network, you map available network ports to a set of logical Ethernet networks. This ensures that the networking environment for each virtual machine is constant regardless of the actual server that it is running on. This allows a virtual machine to be easily migrated between servers within a server pool without affecting the networking for the virtual machine. You perform this mapping in Oracle VM Manager.

In Oracle VM a network can perform one or more network functions. Oracle VM has the following network functions: server management, live migrate, cluster heartbeat, virtual machine, and storage. Functions can be combined or spread over several different networks; this design decision depends on the available physical network infrastructure, such as the number of NICs in each server, network traffic and appliction behaviors within the system.

Jobs are a sequence of operations usually triggered by a user action. For example: discovering a server, presenting a repository, creating a VM, and so on. These jobs appear in the Jobs Summary pane at the bottom of the Oracle VM Manager user interface and their status is refreshed according to their progress. Some jobs are not the result of a specific user action but are a recurring system operation, such as checking the YUM repository for updates. A history of all jobs in the environment is available in the Jobs tab, where you can view and filter the job list and display details of each job: status, execution time stamps, operations executed as part of the job, etc.

Since jobs are sequential and sometimes take time to complete, tracking the status of a job within the Jobs tab allows you to understand what actions the system is currently performing, and which actions are queued to run in sequence after the current job has completed. Jobs also allow you to access system messages that may be useful to debug the failure of an operation.

Events are often also related to user actions, but their main function from a user perspective is to register status information of "objects" for future reference or to make problems easier to trace back. Events are displayed in the Servers and VMs, Repositories, and Storage tabs in the Oracle VM Manager user interface and the list of events depends on the object selected in the tree view of the Navigation pane. For example, the events list of a VM shows you when it was created, at what point it was started and stopped, when it was migrated, and so on. If you select a server or server pool in the same tree view, different types of events appear, related to that particular object. Events have no status but a severity level: most events will be informational, but they can also be warnings, errors or other situations that require your attention. You must acknowledge an error event to clear the error. See Section B.1.11, “Acknowledging Events/Errors” for information on acknowledging events.

Events represent changes in the state of a system that may not be associated with a job. For instance, events can be triggered through changes in the environment such as server crashes or storage disconnects. Therefore, events are used to alert you to potential problems that may need your attention.