How to fix an Unstable Build
For those of us that don't speak build-farm, how do I make this build not unstable?
ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
For those of us that don't speak build-farm, how do I make this build not unstable?
Looking at the console output (http://jenkins.ros.org/job/doc-hydro-navigation/26/console), the only error I see is that it barfs when trying to send emails -- I'm guessing hersh@willowgarage.com (the maintainer email) is no longer valid?
The line "Notification recipients: hersh@wDocumenting a catkin package" will extend the list of recipients and that might break the email notification plugin. The problem here seems to be that the output is mixed up (like stdout and stderr). I am pretty sure I filled a ticket about that some time ago but can't find it.
I'd suggest running rosdoc_lite on your latest source in a catkin workspace.
See if that works, and what it says.
EDIT: to better duplicate the build farm doc environment, use a clean workspace that has not yet been built.
On that page there is a link called "Test Result". If you click on it you'll see that 1/1 tests failed. If you fixup that test it will likely be stable again. You can also get an unstable label if some warnings are detected, though I'm not sure how strict it is.
EDIT:
Well this is a doc job, so there is likely something else going on here (but the above still explains how/why a job is considered "unstable").
Asked: 2013-09-27 05:57:03 -0500
Seen: 332 times
Last updated: Sep 29 '13
Adding postinst to debian release
How is the info in the package.xml of a metapackage used?
My package added to shadow-fixed repository with bloom is installed into a wrong path
buildfarm Krel_import-package checksum error
How to catkin_package to bloom Debian library packages?
Why the ROS 2 Buildfarm package blacklisting is not configurable per-repository or per-package?
How to test the private functions inside a ROS node?