摘要:
A system and techniques are described for managing key-value pairs using an architecture that does not restrict its users to any specific platform or storage subsystem. According to one aspect of the invention, the repository is portable in that the architecture need not change based on the environment and platform in which the repository is used. Rather, the portion of the architecture that is platform-specific is limited to plug-ins at a storage abstraction layer of the repository. The plug-ins expose the same storage abstraction interface to the other layers of the repository, but are implemented to interact with different platforms and storage subsystems. Consequently, in response to being ported from one platform to another, the repository may simply change the plug-in to invoke for persistent storage operations.
摘要:
A system and techniques are described for managing key-value pairs using an architecture that does not restrict its users to any specific platform or storage subsystem. According to one aspect of the invention, the repository is portable in that the architecture need not change based on the environment and platform in which the repository is used. Rather, the portion of the architecture that is platform-specific is limited to plug-ins at a storage abstraction layer of the repository. The plug-ins expose the same storage abstraction interface to the other layers of the repository, but are implemented to interact with different platforms and storage subsystems. Consequently, in response to being ported from one platform to another, the repository may simply change the plug-in to invoke for persistent storage operations.
摘要:
A new instance of an application is added to a group of existing instances (of that application) that share a resource, such as a database. The new instance is added by creating a new object for the new instance by cloning an existing object of an existing instance in the group, setting up connectivity between the new instance and a network through which the multiple instances communicate with one another, and starting up the new instance. The just-described acts are performed in the reverse order when deleting an existing instance. Each of these acts may be performed manually, or automatically without user input, with the user issuing at least one instruction to add or delete an instance.
摘要:
In an application executing in multiple computers connected by a network, an instance of the application is deleted from a group of instances that share a resource, such as a database. Specifically the to-be-deleted instance is shut down, followed by deleting connectivity between the to-be-deleted instance and the network, and then deleting an object of the to-be-deleted instance. Each of these acts is performed automatically without user input, once the user issues an instruction to delete the to-be-deleted instance.
摘要:
Less-restrictive techniques are provided for ensuring that replicated-data systems will never provide out-of-date version of data items. A replicated-data system maintains a version number, a membership group identifier, and a membership count, with each replica of a data item. These values are maintained in such a way as to allow the replicated-data system to reliably satisfy some read requests even though half, or less than half, of the replicas of the data item are available.
摘要:
Allowing for resource attributes that may change dynamically while the resource is in use, provides for dynamic changes to the manner in which such resources are managed. Management of dynamic resource attributes by clusterware involves new entry points to clusterware agent modules, through which resource-specific user-specified instructions for discovering new values for resource attributes, and for performing a user-specified action in response to the new attribute values, are invoked. A clusterware policy manager may know ahead of time that a particular resource has dynamic attributes or may be notified when a resource's dynamic attribute has changed and, periodically or in response to the notification, request that the agent invoke the particular resource-specific instructions for discovering new values for attributes for the particular resource and/or for performing a user-specified action in response to the new attribute values. During the majority of this process, the resource remains available.
摘要:
There are disclosed methods and a packet receiver for testing a network. A packet may be received from the network at a port unit. The port unit may ignore the received packet if the received packet does not contain a valid test signature. When the received packet contains a valid test signature, the port unit may determine if the port unit is an expected destination for the received packet. When the port unit is not an expected destination for the received packet, a misdirected packet counter may be incremented. When the port unit is an expected destination for the received packet, test data may be extracted from the received packet.
摘要:
There are disclosed methods and a packet receiver for testing a network. A packet may be received from the network at a port unit. The port unit may ignore the received packet if the received packet does not contain a valid test signature. When the received packet contains a valid test signature, the port unit may determine if the port unit is an expected destination for the received packet. When the port unit is not an expected destination for the received packet, a misdirected packet counter may be incremented. When the port unit is an expected destination for the received packet, test data may be extracted from the received packet.
摘要:
Techniques are provided for managing replicated data items in a cluster of shared nothing nodes. The techniques include synchronously propagating changes to one or more replicas, while asynchronously propagating changes to the remaining replicas. Because the replicas to which the change is synchronously replicated may be a relatively small percentage of the total number of replicas, the overhead associated with performing an update is reduced. The process of selecting which nodes are involved in the synchronous update operation may take into account a variety of factors, including which replicas are available at the time the update operation is being performed.
摘要:
A system and techniques are described for creating member private data in a global namespace. According to one aspect, first data that is associated with a key is stored in a repository shared by multiple members. After the first data has been stored, second data that is associated with the key is stored without overwriting the first data. Mapping data is stored. The mapping data maps the key, relative to a first member, to a first data structure that stores the first data. The mapping data also maps the key, relative to a second member, to a second data structure that stores the second data. Based on the mapping data, the first member is allowed to access the first data and not the said second data. Based on the mapping data, the second member is allowed to access the second data and not the first data. The first data and the second data may include key values and subtrees of child keys.