Method and apparatus for managing workflow failures by retrying child and parent elements

A method of managing a workflow failure in a distributed computing environment of a computing grid, a workflow having a plurality of workflow elements, the method comprising:

• associating each of the plurality of workflow elements with a retry value that indicates a maximum number of times the corresponding workflow element can be retried, wherein a first retry value is associated with a first element of the workflow;
• and a second retry value is associated with a second element of the workflow;
• identifying, by using a processor, one or more elements of the plurality of workflow elements that do not successfully execute in the distributed computing network of the computing grid, wherein the computing grid comprises grid management for allocating resources among one or more grid components to provide parallel processing of the plurality of workflow elements;
• and
• retrying the one or more identified elements that do not successfully execute a number of times equal to an associated retry value for the one or more identified element, wherein the workflow comprises a hierarchy with at least a parent element having a plurality of child elements;
• in response to the one or more identified elements not successfully executing after retrying the number of times equal to the associated retry value, determining whether the one or more identified elements has a parent element and retrying the parent element according to a retry value associated with the parent element such that retries proceeds up the hierarchy from a child element of the plurality of child elements to the parent element when the child element does not successfully execute after retrying a number of times equal to an associated retry value for the child element.

A method of managing a parent workflow that does not successfully execute in a distributed computing environment of a computing grid, the parent workflow having a plurality of child workflow elements, the method comprising:

• associating each of the plurality of child workflow elements with a retry value that indicates a maximum number of times the corresponding child workflow element can be retried, wherein a first retry value is associated with a first child workflow element of the parent workflow and a second retry value is associated with a second child workflow element of the parent workflow;
• identifying, by using a processor, one or more child workflow elements that do not successfully execute in the distributed computing network of the computing grid, wherein the computing grid comprises grid management for allocating resources among one or more grid components to provide parallel processing of the plurality of workflow elements;
• and
• retrying the identified child workflow elements that did not successfully execute a number of times equal to an associated retry value for the identified child workflow elements, wherein the workflow comprises a hierarchy with at least the parent workflow having the plurality of child workflow elements;
• in response to the identified child workflow elements not successfully executing after retrying the number of times equal to the associated retry value, determining whether the identified child workflow elements has a parent workflow and retrying the parent workflow according to a third retry value associated with the parent workflow such that retries proceed up the hierarchy from a child element of the plurality of child workflow elements to the parent workflow when the child element does not successfully.

An apparatus for managing a workflow failure in a distributed computing environment of a computing grid, comprising:

• a processor programmed for:
• identifying a first element of a workflow that does not successfully execute in the distributed computing network of the computing grid, wherein the computing grid comprises a grid management for allocating resources among one or more grid components to provide parallel processing of a plurality of elements of the workflow, wherein the workflow comprises a hierarchy with at least a parent element having a plurality of child elements, wherein each of the plurality of elements of the workflow is associated with a retry value that indicates a maximum number of times the corresponding element of the workflow can be retried;
• retrying the first element that does not successfully execute a number of times equal to an associated retry value for the first element, wherein the associated retry value is a first retry value;
• and
• In response to the first element not successfully executing after retrying the number of times equal to the first value, determining whether the first element has a parent element and retrying the parent element according to a retry value associated with the parent element such that retries proceed up the hierarchy to a higher parent element when any of the plurality of child elements does not successfully execute.

A computer program product comprising a non-transitory computer-usable storage medium having executable code which, when executed by a processor, causes the processor to execute a process for managing a workflow failure in a distributed computing environment of a computing grid, comprising:

• identifying a first element of a workflow that does not successfully execute in the distributed computing network of the computing grid, wherein the computing grid comprises a grid management for allocating resources among one or more grid components to provide parallel processing of a plurality of elements of the workflow, wherein the workflow comprises a hierarchy with at least a parent element having a plurality of child elements, wherein each of the plurality of elements of the workflow is associated with a retry value that indicates a maximum number of times the corresponding element of the workflow can be retried;
• retrying the first element that does not successfully execute a number of times equal to an associated retry value for the first element, wherein the associated retry value is a first retry value;
• and
• In response to the first element not successfully executing after retrying the number of times equal to the first value, determining whether the first element has a parent element and retrying the parent element according to a retry value associated with the parent element such that retries proceed up the hierarchy to a higher parent element when any of the plurality of child elements does not successfully execute.