org.apache.ignite.spi.failover

Interface FailoverSpi

All Superinterfaces:

IgniteSpi

All Known Implementing Classes:

AlwaysFailoverSpi, JobStealingFailoverSpi, NeverFailoverSpi

public interface FailoverSpi
extends IgniteSpi

Failover SPI provides developer with ability to supply custom logic for handling failed execution of a grid job. Job execution can fail for a number of reasons:

• Job execution threw an exception (runtime, assertion or error)
• Node on which job was execution left topology (crashed or stopped)
• Collision SPI on remote node cancelled a job before it got a chance to execute (job rejection).

In all cases failover SPI takes failed job (as failover context) and list of all grid nodes and provides another node on which the job execution will be retried. It is up to failover SPI to make sure that job is not mapped to the node it failed on. The failed node can be retrieved from GridFailoverContext.getJobResult().node() method.

Ignite comes with the following built-in failover SPI implementations:

• NeverFailoverSpi
• AlwaysFailoverSpi
• JobStealingFailoverSpi

NOTE: this SPI (i.e. methods in this interface) should never be used directly. SPIs provide internal view on the subsystem and is used internally by Ignite kernal. In rare use cases when access to a specific implementation of this SPI is required - an instance of this SPI can be obtained via Ignite.configuration() method to check its configuration properties or call other non-SPI methods. Note again that calling methods from this interface on the obtained instance can lead to undefined behavior and explicitly not supported.

Method Summary

Modifier and Type	Method and Description
ClusterNode	failover(FailoverContext ctx, List<ClusterNode> top) This method is called when method ComputeTask.result(org.apache.ignite.compute.ComputeJobResult, List) returns value ComputeJobResultPolicy.FAILOVER policy indicating that the result of job execution must be failed over.

Methods inherited from interface org.apache.ignite.spi.IgniteSpi

getName, getNodeAttributes, onClientDisconnected, onClientReconnected, onContextDestroyed, onContextInitialized, spiStart, spiStop

Method Detail

failover

ClusterNode failover(FailoverContext ctx,
                     List<ClusterNode> top)

This method is called when method ComputeTask.result(org.apache.ignite.compute.ComputeJobResult, List) returns value ComputeJobResultPolicy.FAILOVER policy indicating that the result of job execution must be failed over. Implementation of this method should examine failover context and choose one of the grid nodes from supplied topology to retry job execution on it. For best performance it is advised that FailoverContext.getBalancedNode(List) method is used to select node for execution of failed job.

Parameters:

ctx - Failover context.

top - Collection of all grid nodes within task topology (may include failed node).

Returns:

New node to route this job to or null if new node cannot be picked. If job failover fails (returns null) the whole task will be failed.