Abstract: Reconfiguring processing groups for cascading data workloads including receiving a request to reconfigure a computing system to execute a workload, wherein the computing system comprises a first processing group and a second processing group, wherein the first processing group comprises a first central processing unit (CPU), a first graphics processing unit (GPU), and a second GPU, and wherein the second processing group comprises a second CPU and a third GPU; reconfiguring the computing system including activating a processor link spanning the first processor group and the second processor group between the second GPU and the third GPU; and executing the workload using the first GPU, second GPU, and third GPU including cascading data, via processor links, from the first CPU to the first GPU, from the first GPU to the second GPU, and from the second GPU to the third GPU.

Abstract: Reconfiguring processing groups for cascading data workloads including receiving a request to reconfigure a computing system to execute a workload, wherein the computing system comprises a first processing group and a second processing group, wherein the first processing group comprises a first central processing unit (CPU), a first graphics processing unit (GPU), and a second GPU, and wherein the second processing group comprises a second CPU and a third GPU; reconfiguring the computing system including activating a processor link spanning the first processor group and the second processor group between the second GPU and the third GPU; and executing the workload using the first GPU, second GPU, and third GPU including cascading data, via processor links, from the first CPU to the first GPU, from the first GPU to the second GPU, and from the second GPU to the third GPU.

Abstract: Reconfiguring processing groups for cascading data workloads including receiving a request to reconfigure a computing system to execute a workload, wherein the computing system comprises a first processing group and a second processing group, wherein the first processing group comprises a first central processing unit (CPU), a first graphics processing unit (GPU), and a second GPU, and wherein the second processing group comprises a second CPU and a third GPU; reconfiguring the computing system including activating a processor link spanning the first processor group and the second processor group between the second GPU and the third GPU; and executing the workload using the first GPU, second GPU, and third GPU including cascading data, via processor links, from the first CPU to the first GPU, from the first GPU to the second GPU, and from the second GPU to the third GPU.

Abstract: Reconfiguring processing groups for cascading data workloads including receiving a request to reconfigure a computing system to execute a workload, wherein the computing system comprises a first processing group and a second processing group, wherein the first processing group comprises a first central processing unit (CPU), a first graphics processing unit (GPU), and a second GPU, and wherein the second processing group comprises a second CPU and a third GPU; reconfiguring the computing system including activating a processor link spanning the first processor group and the second processor group between the second GPU and the third GPU; and executing the workload using the first GPU, second GPU, and third GPU including cascading data, via processor links, from the first CPU to the first GPU, from the first GPU to the second GPU, and from the second GPU to the third GPU.

Abstract: Domains can also be used to control access to physical memory space. Data in a physical memory space that has been used by a process sometimes endures after the process stops using the physical memory space (e.g., the process terminates). In addition, a virtual memory manager may allow processes of different applications to access a same memory space. To prevent exposure of sensitive/confidential data, physical memory spaces can be designated for a specific domain or domains when the physical memory spaces are allocated.

Abstract: When an operating system process evaluates a rule for an operation being attempted on a logical network port, the operating system process determines whether the target logical port falls within a range of logical ports, and then determines whether the operation is associated with a permitted domain of the range of logical ports. If the operation is a bind operation, then the process attempting to bind to the target port will be allowed to bind if the target port falls within the range and the operation/process is associated with a permitted domain. Otherwise, the binding operation will not be allowed to proceed.

Abstract: Domains can be used to secure resources of a cluster. An administrator can configure a node of a cluster as a member of a particular domain. Membership in a cluster can be restricted to nodes that are members of the particular domain. When a node generates a cluster message, a kernel process or operating system process of the node will indicate the domain(s) of the node in the cluster message. The cluster message can be a command message to read or write to a storage resource of the cluster. When the cluster storage resource node or node that controls the storage resource receives the command message, the node will examine the command message to ensure the message indicates a domain that aligns with the cluster. If the proper domain is indicated in the command message, then the command message is processed. Otherwise, the command message is denied.

Abstract: Functionality can be implemented in an operating system to increase the granularity of isolation for objects. A domain can be defined to represent each of different entities (e.g., different departments or work groups). User identifiers and/or user credentials can be associated with the appropriate domain or domains. An administrator can then define a set of rules that govern operation(s) that can be performed on the objects based on the domains. Processes running on a system will inherit the domains of a user account logged into the system. When a process running on the system attempts to perform an operation on an object, an operating system process evaluates the domain isolation rules with an identifier of the object and a domain identifier to determine whether the operation is permitted to proceed.

Abstract: Domains can be used to secure resources of a cluster. An administrator can configure a node of a cluster as a member of a particular domain. Membership in a cluster can be restricted to nodes that are members of the particular domain. When a node generates a cluster message, a kernel process or operating system process of the node will indicate the domain(s) of the node in the cluster message. The cluster message can be a command message to read or write to a storage resource of the cluster. When the cluster storage resource node or node that controls the storage resource receives the command message, the node will examine the command message to ensure the message indicates a domain that aligns with the cluster. If the proper domain is indicated in the command message, then the command message is processed. Otherwise, the command message is denied.

Abstract: Domains can also be used to control access to physical memory space. Data in a physical memory space that has been used by a process sometimes endures after the process stops using the physical memory space (e.g., the process terminates). In addition, a virtual memory manager may allow processes of different applications to access a same memory space. To prevent exposure of sensitive/confidential data, physical memory spaces can be designated for a specific domain or domains when the physical memory spaces are allocated.

Abstract: Functionality can be implemented in an operating system to increase the granularity of isolation for objects. A domain can be defined to represent each of different entities (e.g., different departments or work groups). User identifiers and/or user credentials can be associated with the appropriate domain or domains. An administrator can then define a set of rules that govern operation(s) that can be performed on the objects based on the domains. Processes running on a system will inherit the domains of a user account logged into the system. When a process running on the system attempts to perform an operation on an object, an operating system process evaluates the domain isolation rules with an identifier of the object and a domain identifier to determine whether the operation is permitted to proceed.

Abstract: When an operating system process evaluates a rule for an operation being attempted on a logical network port, the operating system process determines whether the target logical port falls within a range of logical ports, and then determines whether the operation is associated with a permitted domain of the range of logical ports. If the operation is a bind operation, then the process attempting to bind to the target port will be allowed to bind if the target port falls within the range and the operation/process is associated with a permitted domain. Otherwise, the binding operation will not be allowed to proceed.

Abstract: A data processing system, method, and product are disclosed for automatically configuring IP security tunnels. A security policy specification format is established that is capable of being utilized by any one of multiple different operating systems and any one of multiple different machine types. An IP security tunnel is automatically configured utilizing the security policy specification format.

Abstract: A system and method for dynamically determining a CRL location and protocol. CRL location names and protocols are retrieved from a digital certificate data structure which includes a network servers that contain the CRL file. A determination is made as to whether any of the servers reside in the current domain, in which case the server is used because the data is more secure. If no locations are within the current domain, Internet servers outside the current domain are analyzed. Security parameters may be established that restrict which Internet servers can be used to retrieve the data. The security parameters may also include which access methods may be used to retrieve data since some access methods provide greater security than other access methods. A security parameter may also be based upon both the access method and the name, or address, of the Internet server.

Abstract: A system and method for providing multiple virtual private networks from a computer system. The computer system communicates with a remote computer system in order to allow encrypted data traffic to flow between the respective systems. Two phases are used to authenticate the computer systems to one another. During the first phase, digital certificates or pre-shared keys are used to authenticate the computer systems. A phase 1 ID rules list contains authentication rules for local-remote computer pairs. During the second phase, a hash value is used to authenticate the computer systems and a security association payload is created. The remote system's IP address is used for connecting. The phase 1 ID rules list corresponds to one or more phase 2 ID rules lists. If the remote ID is not found in the phase 2 ID rules list, a default rule is used based upon the phase 1 ID rules list.

Abstract: A system and method for managing multiple proxy servers by a client computer. In one embodiment, the client computer's configuration is set to the fastest proxy server available. In another embodiment, a proxy table is accessed by the client computer to determine which proxy to use for a given web address. When a web address is included in the table, the corresponding proxy server is used to request the contents of the web address. If the web address is not included in the table, a default proxy server is used to request the information. In another embodiment, a periodic test is made to determine the speed of the current proxy server. If the speed is less than a predetermined threshold, the available proxy servers are all tested and the best-performing proxy server is selected. In another embodiment, aspects of each of the above-described embodiments are combined to provide a proxy server to a client computer based upon either a given web address (URL) or the fastest current proxy server.

Abstract: A system and method for dynamically integrating a site specific search engine. In one embodiment, a hidden HTML tag is used on web pages to identify the site specific search engine for the web site. When a user requests a site specific search engine from the user's web browser, the HTML tag is located and the corresponding web page containing the site specific search engine is opened. In another embodiment, if the HTML tag identifying the site specific search engine is not found, an analysis of the web site's root node is made to identify a site specific search engine. If a site specific search engine is identified through the analysis, the search engine web page is opened. In yet another embodiment, if the site specific search engine is not found through either a hidden HTML tag or by analyzing the web site's root node, a generic site specific search routine is invoked.

Abstract: In a search session through a browser for pages on the World Wide Web (Web), a system for precluding repetitive accessing of documents linked to hyperlinks in a plurality of hypertext documents. The system is directed to a Web communication network with user access via a plurality of data processor controlled interactive receiving display stations for displaying received hypertext documents of at least one display page containing text, images and a plurality of embedded hyperlinks, each hyperlink being user selectable to access and display a respective linked hypertext document. In addition to precluding hyperlinks which had been activated in previous pages and/or hyperlinks which have been discounted merely by having been present in previous pages, the Web browser may be setup to permit the user to specifically discount individual hyperlinks or groups of hyperlinks. Also, the Web browser may be setup to preclude repetitive hyperlinks in subsequent Web pages from searches from different search engines.

Abstract: A data processing system, method, and product are disclosed for automatically configuring IP security tunnels. A security policy specification format is established that is capable of being utilized by any one of multiple different operating systems and any one of multiple different machine types. An IP security tunnel is automatically configured utilizing the security policy specification format.

Abstract: A system and method for dynamically determining a CRL location and protocol. CRL location names and protocols are retrieved from a digital certificate data structure which includes a network servers that contain the CRL file. A determination is made as to whether any of the servers reside in the current domain, in which case the server is used because the data is more secure. If no locations are within the current domain, Internet servers outside the current domain are analyzed. Security parameters may be established that restrict which Internet servers can be used to retrieve the data. The security parameters may also include which access methods may be used to retrieve data since some access methods provide greater security than other access methods. A security parameter may also be based upon both the access method and the name, or address, of the Internet server.