摘要:
Provided are a method, system, program and device for maintaining shadow page tables in a sequestered memory region. A first processor executing an application invokes a second processor to create a shadow page table used for address translation for the application in a sequestered memory region non-alterable by processes controlled by an operating system executed by the first processor. The shadow page table references at least one page in an operating system memory region accessible to processes controlled by the operating system.
摘要:
Systems and methods are provided through which a packet is transferred within a router through the use of a switch-label. The switch-label is generated upon receipt by a component of the router. The switch label indicates an address within the router of another component within the router that will transmit the packet to a destination outside the router. The switch-label uses an abbreviated address of components within the router. The address is abbreviated because the number of potential addresses within a router is significantly less than the number of addresses outside the router that require a fill address. As a result, transferring of the packet within the router is faster because processing the switch-label using an abbreviated address is faster than processing a full address.
摘要:
A network device may be configured with a primary control element and a second control element. Both the primary and secondary control elements may receive information from one or more forwarding elements. Both the primary and secondary control elements may send control information to the forwarding element. The control information from the secondary control element, however, may be discarded either at the forwarding element or at a blocking agent between the secondary control element and the forwarding element. In this manner, the secondary control element may be synchronized with the primary control element in a stateless manner. If a failure condition is detected, control plane operations may be performed using the secondary control element. As a result, the network device may become fault tolerant, and experience reduced amounts of down time due to failure of one or more elements within the network device.