| Summary: | discuss mapping for Koji status codes | ||
|---|---|---|---|
| Product: | z_Archived | Reporter: | Ken Leung <plcnkl> |
| Component: | Mylyn | Assignee: | Mylyn Inbox <mylyn-inbox> |
| Status: | CLOSED WORKSFORME | QA Contact: | |
| Severity: | enhancement | ||
| Priority: | P3 | CC: | plcnkl, swagiaal |
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Linux | ||
| Whiteboard: | |||
|
Description
Ken Leung
The framework defines the following states:
BuildState { QUEUED, RUNNING, STOPPED, BUILDABLE; }
BuildStatus { FAILED, UNSTABLE, SUCCESS, DISABLED, ABORTED; }
Can you suggest a mapping and let me know which states or statuses are missing? It seems to me that we would be able to represent the Koji states with the existing API fairly well.
Hi Mr. Pingel: In this case, since I have access to both the build and task data, I can combine those koji status codes and map to the existing mylyn builds status code. So the proposed mapping plan is: mylyn builds koji build states task status QUEUED Assigned RUNNING Open STOPPED BUILDABLE Free mylyn builds koji build status task status FAILED Failed UNSTABLE - SUCCESS Closed DISABLED Cancelled ABORTED Notice that there are a few missing ones, I guess I need more understanding about Hudson's stopped and aborted status. Are they caused by a build fail event? Or are they user triggered (like cancel build)? For the stopped status, my guess is I can set it as stopped if the build status is not null (in this case, when the build is running, should the build status be set as null?) , am I correct? (In reply to comment #2) > Notice that there are a few missing ones, I guess I need more understanding > about Hudson's stopped and aborted status. Are they caused by a build fail > event? Or are they user triggered (like cancel build)? STOPPED means that a build has completed regardless of its result. DISABLED means that a plan (job) is disabled and no builds will run. This state doesn't make much sense for builds. ABORTED indicates that a build was canceled by a user. > For the stopped status, my guess is I can set it as stopped if the build status > is not null (in this case, when the build is running, should the build status be > set as null?) , am I correct? Yes, I believe that is correct. I'll mark this bug as resolved for now. Please feel free to reopen if this needs more discussion or file specific bug reports if the existing status codes should be extended. |