Story (Required)

As a person reviewing CI failures, I should be able to evaluate what happened from just the pipeline run log.

Background (Required)

Many tests create a pipeline run on testing cluster and expect this pipeline run to pass. When pipeline run on testing cluster fails, the test suit fails which in turn fails the pipeline run on CI clusters and we get a Slack notification containing link to logs.

However, these logs only contain a message saying something like

Failed Step: Start and verify pipeline "s2i-nodejs-pipeline" with param "VERSION" with values stored in variable "nodejs-tags" with workspace "name=source,claimName=shared-pvc"
        Specification: specs/clustertasks/clustertask-s2i.spec:92
        Error Message: error waiting for pipeline run s2i-nodejs-pipeline-run-62dvh to finish 
         "s2i-nodejs-pipeline-run-62dvh" failed

which is not enough to know what broke. The engineer than needs to log in to the cluster, find the pipeline run and investigate logs and/or events.

Moreover, even if we classify the failure correctly in Report Portal, the system will not be able to automatically classify similar failures because of lack of this additional context.

Acceptance Criteria (Mandatory)

Logs in failed CI pipeline run contain additional information so that it's not necessary to log in to testing cluster in most cases.