Module datarush.commons
Package com.pervasive.datarush.cal
package com.pervasive.datarush.cal
Provides interfaces the define the "cluster abstraction layer".
NOTE: the interfaces in this package should be considered
unstable.
The cluster abstraction layer allows the DataRush engine to be implemented on
any cluster. In order to provide an implementation for a new type of cluster
one needs to register a new
ClusterProvider and, optionally, a
StorageManagerProvider.-
ClassDescriptionA session qualifier to provide uniqueness of session attributes.Subclass of DistributedExecutionException that contains all the exceptions that occurred in one batch.Enumeration of possible caches for
Element's.The classpath specifies for a job.An element of a classpath.Extension point to register cluster providers for new schemes.Facade to lookup providers from META-INF/services.A "URL-like" entity that identifies a cluster.A distributed equivalent of aCallable.A distributed equivalent of aCallable.Distributed context passed toDistributedCallable.init(DistributedCallableContext).Distributed context passed toDistributedCallableBatch.init(DistributedCallableContextBatch).Convenient base class forDistributedCallable's that do not support monitoring/partial results.Convenient base class forDistributedCallableBatch's that do not support monitoring/partial results.AnExecutionExceptionin a distributed context.Distributed equivalent of theExecutorServiceconcept.Miscellaneous execution options.Miscellaneous utilities related toDistributedExecutorServiceand its subordinate classes.JobCallable<T>Context available to job.Client-side API for submitting units of work to the job master.Factory for creating newJobExecutorService's.Basic job information that contains the job name, id, and the user that submitted the job.Rollup of all parameters needed to launch a jobExtension ofCallablethat provides a way to get its current status prior to execution completion.Extension ofFuturethat allows the client to query for the partial status.Represents the initial node allocation.Assignment details for a particular node ID.Reflects the current state of node allocation, which takes into account failed nodes.Assignment details for a particular node ID.Contains node and host informationContains host information associated with the nodes in the cluster that the current job is running on.Extra options for an OSGI classpath elementInformation pertaining to an instance of a partitioning operation.Subinterface of DistributedExecutorService for running in pseudo-distributed mode.Describes the resource allocation for a job server process.Resource specifier for the resources needed by a job.ResultMerger<R>Provides a key for a session attribute and, optionally, an initial valueConvenient pairing ofStorageBucketIdand it's associatedStorageManager.A storage bucket represents one unit of work to be exchanged by nodes in the cluster.A manager of intermediate storage.Extension point to register a newStorageManagertype.Utilities for creating storage managersRollup of the various parameters needed to create aStorageManager.An allocator ofStorageBucketId's.Identifies a "virtual node" in a cluster.Enumeration of non-parallel locations