Abstract: A decentralised computer storage network is described comprising a plurality of hypervisors or nodes HV1, HV2, HVn. Each storage region in a node is controlled by an associated software controller 11, 12. Each node is configured to prepare a compressed metadata file detailing a snapshot of the data stored in its storage drive. The nodes can broadcast their metadata files so that they can be received by other nodes in the network. This allows each node to be aware of the data stored in the other nodes. By maintaining a local record of metadata files received over the network, each storage node can establish a live picture of the data stored in other storage nodes. This can permit any individual storage node to take decisions that control the distribution of data in the network.
Abstract: A CDN accelerator computer system (32) is described that can allow a website to be provided in a content delivery network (CDN). The CDN accelerator (32) is configured to receive a request for web content from a user (20). Processors in the CDN accelerator can then retrieve HTML source code from an origin server (34) and modify the source code by changing the addresses of static content items to refer to addresses in a CDN. The CDN accelerator (32) can then send the modified HTML source code to the user (20), and the user's browser can acquire the static content items from the CDN.
Abstract: A decentralised computer storage network is described comprising a plurality of hypervisors or nodes HV1, HV2, HVn. Each storage region in a node is controlled by an associated software controller 11, 12. Each node is configured to prepare a compressed metadata file detailing a snapshot of the data stored in its storage drive. The nodes can broadcast their metadata files so that they can be received by other nodes in the network. This allows each node to be aware of the data stored in the other nodes. By maintaining a local record of metadata files received over the network, each storage node can establish a live picture of the data stored in other storage nodes. This can permit any individual storage node to take decisions that control the distribution of data in the network.