[wip] make resource init failures result in step retries

[OwenKephart]

Summary & Motivation

this does not work at all, just trying to narrow in on a potential solution

How I Tested These Changes

Changelog

Insert changelog entry or delete this section.

[OwenKephart]

• [wip] make resource init failures result in step retries #29619 👈 (View in Graphite)
• master

This stack of pull requests is managed by Graphite. Learn more about stacking.

[graphite-app]

The empty SerializableErrorInfo("", [], "") will cause issues with the retry mechanism. The get_failure_or_retry_event_after_crash method relies on this error information to determine retry behavior and generate appropriate failure reports.

Consider extracting the actual error details from the resource initialization failure event instead. This would ensure the retry logic has complete information to make correct decisions about whether to retry the step and how to report the failure to users.

# Extract error info from the resource init failure event
error_info = event_or_none.event_specific_data.error

Suggested change

	SerializableErrorInfo("", [], ""),
	active_execution.get_known_state(),
	event_or_none.event_specific_data.error,
	active_execution.get_known_state(),

Spotted by Diamond

Is this helpful? React 👍 or 👎 to let us know.