US 20050083960 Method and apparatus for transporting parcels of data using network elements with network element storage
A network element with network element storage and independent intelligence may be configured to provide temporary mass storage to facilitate the transfer of large files across an optical network. The network element may also be provided with intelligence to enable the network element to maintain a higher level understanding of the data flows. Using network element storage enables network elements involved in data transmission across the network to temporarily store data being transferred on the network. This allows parcels of data to be transmitted part way through the network when a complete path through the network is not available. It also allows data to be aggregated at strategic locations on the network, such as at the location of a transmission bandwidth mismatch, to enable the data to be transmitted over the high capacity optical resource at a higher rate, thus more efficiently utilizing the bandwidth on the higher bandwidth resource.
1. Field of the Invention
This application relates to communication networks and, more particularly, to a method and apparatus for transporting parcels of data using network elements with network element storage.
2. Description of the Related Art
Data communication networks may include various computers, servers, nodes, routers, switches, hubs, proxies, and other devices coupled to and configured to pass data to one another. These devices will be referred to herein as “network elements.” Data is communicated through the data communication network by passing protocol data units, such as frames, packets, cells, or segments, between the network elements by utilizing communication links formed according to a conventional technology, such as optical, electrical, or wireless technology. A particular protocol data unit may be handled by multiple network elements and cross multiple communication links as it travels between its source and its destination over the network.
Grid networks is an emerging application that builds overlay networks, i.e. computational Grids, on existing network infrastructures using Grid computing technology. In a Grid network, which forms a virtual organization, Grid nodes are distributed widely and share computational resources such as disc storage, storage servers, shared memory, computer clusters, data mining, and visualization centers, although other resources may be available as well. One example of Grids is the TeraGrid, in which Grid computing technology has been deployed to enable supercomputer clusters distributed in four distant locations in the United States to collaboratively work on computationally intense tasks, such as high-energy physical simulations and long-term global weather forecasting. Other potential uses for Grid computing include genomics, protein structure research, computational fluid dynamics, astronomy and astrophysics, Search for ExtraTerrestrial Intelligence (SETI), computational chemistry, “intelligent” drug design, electronic design automation, nuclear physics, and high-energy physics. Grid computing may be used for many other purposes as well, and this list is not intended to be inclusive of all possible uses.
Some of these applications are capable of producing an incredible amount of data that must be distributed to other Grid applications for analysis. For example, high energy physics may generate more than a petabyte of data (1 petabyte=1000 Terabyte=1015 bytes). To transfer a petabyte of data over a 10 Gigabit link would take approximately 27.8 hours, assuming 100% throughput, no overhead associated with packet headers, etc., and no network problems. This data must be sent to research facilities and universities around the world for analysis and storage.
When faced with data volumes this large, and data transfer rates this fast, traditional packet switched networks, such as TCP/IP based communication networks, tend to become overloaded and incapable of or inefficient at handling these large data transfers. One technology that is capable of handling these large data transfers is the use of optical networking, which can handle data transfer rates exceeding 10 gigabits per second (Gbps).
There are several interfaces between the optical network and other portions of the network at which there may be a potential transmission rate mismatch. For example, standard computer equipment may not be capable of outputting data at rates as fast as optical networking resources are able to transmit it. Similarly, a bandwidth mismatch may occur where optical networking technology interfaces with other portions of the communication network. For example, assume a network application is interfaced with a 1 Gbps link to an optical network, at which point the traffic is placed onto a 10 Gbps lambda. In a packet-based multiplexed network this is not a problem, since the other 9 Gbps may be occupied by other flows so that the network resource is able to be largely utilized. Where the lambda is a switched resource that has been reserved to carry traffic only for that network application to guarantee its availability, the lower bandwidth feed link acts as a bottleneck that prevents the full capacity of the 10 Gbps lambda from being utilized. This situation is prevalent, for example, where the reserved network bandwidth is a wavelength in a DWDM fiber configured to operate at an OC-192 data transmission rate (10 Gbps). Thus, a large amount of bandwidth may be wasted due to transmission mismatches between the packet network and the optical underlay network.
Additionally, in some instances, it is desirable to send a given large quantity of data to more than one intended recipient. Where the data transmission is performed in a point-to-point manner, this requires the data source to be available for multiple transmission sessions, thus consuming additional data transmission resources.