Details
-
New Feature
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Done
-
-
Pipeline - October, Pipeline - December
Description
As a user of pipeline, I WANNA GO FAST.
To do that, I'm okay with sacrificing some of the ability to resume a pipeline after a hard Jenkins crash, as long as it means pipelines run fast. It would be nice if I can still resume the pipeline after a clean/safe restart of the master, though!
This differs from https://issues.jenkins-ci.org/browse/JENKINS-30896 because for this we're sacrificing some of the guarantees pipeline normally provides.
Concrete needs:
- Implement storage engine (workflow-support)
- Implement options to select that storage engine (workflow-job)
- Implement ability to RUN that storage engine (workflow-cps)
Attachments
Issue Links
- is blocked by
-
JENKINS-47172 Provide more granular control of pipeline flownode persistence
-
- Closed
-
- is related to
-
JENKINS-30896 Unoptimized node storage
-
- Closed
-
-
JENKINS-33761 Ability to disable Pipeline durability and "resume" build.
-
- Closed
-
- relates to
-
JENKINS-49086 Lazy-initialize FlowNodeStorage and heads, etc
-
- Closed
-
basil Thank you for reporting your issues, and I'm sorry to hear they bit you rather hard. I've been doing some investigation on this – it sounds like you may be hitting a pre-existing bug that we saw reported couldn't reproduce (very similar symptoms around use of nesting blocks and executors). However whether it's pre-existing or a regression from this, if a nested pair of nodes enables us to reproduce (and thus solve it), then you can look forward to a fix soonishly.