Tags: titirigaiulian/echo
Tags
feat(trigger): adds pipeline trigger status field (spinnaker#289)
fix(fiat): Set `@EnableFiatAutoConfig` to avoid unnecessary auth requ… …irements (spinnaker#286) Original intention was to not require full blown `fiat` to use `FiatStatus` but unfortunately spring security is brought in as a transitive dependency. This PR is a simpler alternative to moving `FiatStatus` and the related configuration properties classes to a new artifact that would not bring in any unnecessary dependencies. fixes spinnaker/spinnaker#2949
chore(*): Bump spinnaker-dependencies to 0.158.0 (spinnaker#280)
fix(pubsub/google): don't restart when subscription doesn't exist (sp… …innaker#276) (spinnaker#279)
fix(cron): fallback for cron triggers with null id (spinnaker#270) * fix(cron): fallback for cron triggers with null id If pipeline configurations with null cron trigger ids somehow went past front50, we would have "flagging" registrations: new triggers would be found, then on the next pass of the PipelineConfigsPollingAgent we would find scheduled actions with no corresponding triggers (since they don't have) a matching id. We would then deregister then, only to register them again right away, and so on. With this change, we no longer need a persisted id for cron triggers in the pipeline configurations (which is bound to be messed with, deleted, copied, mangled...). This essentially deprecates the stored trigger ids in pipeline configs in favor of the dynamically generated ones generateFallbackId() hashes the toString() representation of Trigger, which should work fine as long as it includes at least the parent pipeline id and the cron expression
fix(metrics): replace Spring Boot metrics with Spectator (spinnaker#267) Also: * pick up new spinnaker-dependencies for spectator 0.68.0 update * update jacoco, which otherwise conflicts with spectator * add explicit dependency on spectatorApi when we use it * add exception tag on error metrics * don't use a mock Registry when unnecessary * make triggers.cronMisfires a timer (so that we can estimate trigger lag) * adjust some log levels and messages * add a pipelineConfigsPollingAgent.executionTimeMillis timer
PreviousNext