Tags: QuanZhang-William/pipeline
Tags
Doubling the MaximumBundleObjects from 10 to 20. The official hard limit is 255 ( for oci images ), and while testing use cases for bundles it was noticed the current max is pretty low, and doubling should still ensure very small oci sizes as well as give a little more flexablitliy in testing. Also updated the tests, so in the future if a change to the max happens, the tests should be g2g with no changes.
[PATCH] Patch knative/pkg to fix HA via StatefulSet Tekton uses a version of knative/pkg in between a regression on the HA via StatefulSet and its fix. Cherry-pick the patch for a minor release. We shall pick-up a newer version of knative/pkg for the next major release instead. Original Knative PR: knative/pkg#2483 Tekton Issue: tektoncd#3404 Signed-off-by: Andrea Frittoli <andrea.frittoli@uk.ibm.com>
Update tutorial links We recently updated the introductory tutorial on the documnentation website. That tutorial covers the same content as the one here. To avoid duplicated efforts and content drift, I'm linkin that doc here and replacing the existing content. Additionally, this removes references to `PipelineResources`.
PipelineRunState cleanup: tests + comments This commit cleans up PipelineRunState logic in preparation for a subsequent commit affecting retries. It adds tests for DAGExecutionQueue and ensures that that function does not return any results if the PipelineRun is cancelled, since we do not want to continue to schedule tasks if the PipelineRun is cancelled. (This is already documented and implemented elsewhere.) It also fixes comments to correctly describe the behavior of GetFinalTasks No functional changes.
PipelineRunState cleanup: tests + comments This commit cleans up PipelineRunState logic in preparation for a subsequent commit affecting retries. It adds tests for DAGExecutionQueue and ensures that that function does not return any results if the PipelineRun is cancelled, since we do not want to continue to schedule tasks if the PipelineRun is cancelled. (This is already documented and implemented elsewhere.) It also fixes comments to correctly describe the behavior of GetFinalTasks No functional changes.
PreviousNext