feat(queue): provide a way for pipeline stages to specify a backoff period (#4841)

[christosarvanitis]

• feat(queue): provide a way for pipeline stages to specify a backoff period

via a new backoffPeriodMs stage configuration property. Before this, pipeline authors had no control over the backoff period. It came from either spinnaker configuration properties or implementations of RetryableTask.getDynamicBackoffPeriod.

This makes it possible for pipeline authors to, for example, control the delay between attempts when webhook stages retry. The actual backoff used is the largest of:

• backoffPeriodMs in the stage
• tasks.global.backOffPeriod
• tasks..backOffPeriod
• tasks...backOffPeriod

• refactor(core): remove duplication in StageExecutionImpl

since getTimeout and getBackoffPeriod were basically the same.

• refactor(core): simplify StageExecutionImpl.getLongFromContext

since Integer, Long and Double all extend from Number. Also, Optional.map isn't necessary since instanceof returns false when value is null.

• docs(queue): update javadoc for RetryableTask.retryableBackOffPeriod

to mention backoffPeriodMs stage property

We prefer small, well tested pull requests.

Please refer to Contributing to Spinnaker.

When filling out a pull request, please consider the following:

• Follow the commit message conventions found here.
• Provide a descriptive summary for your changes.
• If it fixes a bug or resolves a feature request, be sure to link to that issue.
• Add inline code comments to changes that might not be obvious.
• Squash your commits as you keep adding changes.
• Add a comment to @spinnaker/reviewers for review if your issue has been outstanding for more than 3 days.

Note that we are unlikely to accept pull requests that add features without prior discussion. The best way to propose a feature is to open an issue first and discuss your ideas there before implementing them.