 |
» |
|
|
|
The vPars (Virtual Partitions) product allows you to run multiple instances
of HP-UX simultaneously on one hard partition[1] by
dividing the hard partition further into virtual partitions.
Each virtual partition is assigned its own subset of hardware, runs
a separate instance of HP-UX, and hosts its own set of applications.
Because each HP-UX instance is isolated from all other instances,
vPars provides application and OS (Operating System) fault isolation.
Each instance of HP-UX can be of a different release, have different
patches, and have a different kernel. | | | |  | NOTE: Virtual Partitions,
nPartitions, and Hard Partitions DefinedIn this document, we have redefined the terms virtual partitions, nPartitions,
and hard partitions: A complex is
the entire partitionable server, including both cabinets, all cells,
I/O chassis, cables, and power and utility components. A cabinet is
the Superdome's hardware "box", which contains the cells, Guardian
Service Processor (GSP), internal I/O chassis, I/O fans, cabinet fans,
and power supplies. A complex has up to two cabinets. A hard
partition is any isolated hardware environment, such
as a nPartition within a Superdome complex or an entire rp7400/N4000 server. A nPartition is
a subset of a complex that divides the complex into groups of cell
boards where each group operates independently of other groups.
A nPartition can run a single instance of HP-UX or be further divided
into virtual partitions. A virtual
partition is a software partition of a hard partition
where each virtual partition contains an instance of HP-UX. Though
a hard partition can contain multiple virtual partitions, the inverse
is not true. A virtual partition cannot span a hard partition boundary. | | | | |
Product
Features | |
A
single hard partition can be divided into multiple virtual partitions. Each virtual partition runs
its own instance of HP-UX. Thus, different applications or multiple
instances of the same application can run in different virtual partitions
on the same hard partition at the same time without conflicts. Each virtual partition is
assigned its own resources (CPU, memory, and I/O), so there are
no resource conflicts between virtual partitions. Virtual partitions can be
of different operating system releases and patch levels. Virtual partitions can be
individually reconfigured and rebooted (for patches and other changes
that require a reboot). Users on one virtual partition
cannot access files or file systems on other partitions (unless
the file systems are NFS-mounted or access is otherwise given through
networking or for cluster-aware volume groups used within MC/ServiceGuard).
Further, users configured on one virtual partition does not imply
a presence on any other partition. Software-related kernel panics
[2],
resource exhaustion failures, and subsequent reboots in one virtual
partition do not affect any other virtual partition. CPUs available at boot time can be added to or removed
from a virtual partition without rebooting.
Why
Use vPars? | |
The following explains some of the advantages of using vPars: vPars
Increases Server Utilization and Isolates OS and Application FaultsIn certain environments one entire server is dedicated to
a single application. When the demand for that application is not
at peak, such as during non-business hours, the server is underutilized.
If many servers are configured this way, you have many servers that
are being underutilized. You can minimize investment and operational
costs by consolidating servers and running multiple applications
on one server; however, this leaves all applications vulnerable
to problems if any one application or their now single OS has problems. vPars provides a software-based solution that supports isolating
OSs and their applications within virtual partitions; thus, OS or
application problems in one virtual partition do not affect OSs
or applications running in other partitions. vPars also allows consolidation of underutilized servers into
one faster server where applications may not be permitted to affect
one another, such as in the case of an ISP running many small e-services
application servers. vPars
Provides Flexibility Through Multiple but Independent OS InstancesvPars offers flexibility by allowing different HP-UX instances,
versions, and patch levels to run on the same server. If you have applications servers that are running different
OS versions, with vPars you can consolidate the application servers
into different virtual partitions on one server, with each virtual
partition running its own OS version. vPars
Provides Flexibility Through Dynamic CPU allocationvPars allows you to reassign CPUs from one virtual partition
to another without rebooting. Two virtual partitions that have different CPU utilization
peak times can have processors moved between them. For example,
a transaction server used primarily during business hours could
have floating CPUs reassigned overnight to a report server. Such
reassignments can be automated, for example, via cron. Because vPars assigns specific hardware resources to specific
virtual partitions, a user on the transaction server at night is
not affected by the reports server's huge consumption of
processing power. A virtual partition uses only the CPUs that you
assign to it; CPUs are not time-sliced across the virtual partitions. |
Printable version
