Protocol Independent Multicast (PIM) Basics
Protocol-Independent Multicast (PIM) is a routing protocol like RIP or OSPF. It was designed to allow multicast routing without needing to rely on other specific unicast routing protocols. There are two main modes in which PIM will operate in two main modes: Dense mode and Sparse mode.
It is possible for a router to manage traffic for groups that are separately using either of these modes:
When functioning in Dense mode, the router assumes that all other routers want to receive multicast data for a specific group. If a router receives these packets and does not need them because there are no clients that it is aware of, it sends a prune message (a message to remove a path) back to the originating router.
In this Dense mode, the initial assumption of the protocol is that all routers want to get the messages or data, so more initial data is sent out over the network. This condition eventually creates a source-based multicast distribution tree.
When running in Dense mode, the prune messages time out every three minutes, at which time group messages are flooded out of all interfaces again until new prune messages are received.
Sparse mode, by contrast, involves the router assuming that no other routers want to get multicast traffic. In this case, the router closest to the host receives a multicast PIM join message from the receiver device.
Directly connected routers send the PIM join message to upstream routers between themselves and the Rendezvous Point (RP), which is a router that is a designated meeting point. The RP’s job is to keep track of all multicast groups in use on the network. The RP then sends join messages upstream toward the source host.
In this case, the fewest possible routers are involved with seeing and handling multicast traffic, which makes them sparse or sparsely used. The figure shows both these operational modes, which may give you an idea about how you might use them.

Cloud Computing Glossary
cloud computing
A networking solution in which everything — from computing power to computing infrastructure, applications, business processes to personal collaboration — is delivered as a service wherever and whenever you need.

Cloud Computing Glossary
cloud service
The delivery of software, infrastructure, or storage that has been packaged so it can be automated and delivered to customers in a consistent and repeatable manner.

Cloud Computing Glossary
deprovision
The release of cloud services that are no longer needed.

Cloud Computing Glossary
federating
Linking distributed resources together over the cloud.

Cloud Computing Glossary
hypervisor
An operating system that acts as a traffic cop, managing the various virtualization tasks in the cloud to ensure that they make things happen in an orderly manner.

Cloud Computing Glossary
multi-tenancy
The sharing of underlying resources by multiple companies over a cloud.

Cloud Computing Glossary
network attached store
Storage that has its own network address through which it is accessed by the network's workstation users. Acronym: NAS

Cloud Computing Glossary
service level agreement
A contract that stipulates the type of service you need from providers and what type of penalties would result from an unexpected business interruption. Acronym: SLA

Cloud Computing Glossary
solution stack
An integrated set of software that provides everything a developer needs to build an application.

Cloud Computing Glossary
storage area network
A storage systems that is flexible and scalable because it's available to multiple hosts at the same time. Acronym: SAN

Cloud Computing Glossary
vertical industry groups
Workgroups comprised of members from a particular industry such as technology and retail.

Cloud Computing Glossary
virtual memory
The portion of your hard drive that Windows uses to expand the available RAM

Cloud Computing Glossary
virtualization
Using computer resources to imitate other computer resources or whole computers to maximize performance and flexibility.