title | linkTitle | weight | description |
---|---|---|---|
Next Release Preview |
Next Release Preview |
2 |
Please make a pull request to describe any changes you wish to highlight in the next release of Spinnaker. These notes will be prepended to the release changelog.
This feature is now enabled by default.
spinnaker/echo#1292 adds a new configuration flag: pipelineCache.filterFront50Pipelines
that defaults to false. When false, Echo caches all pipelines Front50. When true, it only caches enabled pipelines with enabled triggers of specific types -- the types that echo knows how to trigger, along with some changes to the logic for handling manual executions so they continue to function. This is typically a very small subset of all pipelines.
spinnaker/orca#4448 adds a new configuration flag: front50.useTriggeredByEndpoint
that defaults to false. When false, Orca queries Front50 for all pipelines each time a pipeline execution completes. When true, Orca only queries for pipelines triggered when a specific pipeline completes which is potentially a very small subset of all pipelines.
spinnaker/clouddriver#5949 changes the validation Clouddriver runs before performing operations for the Kubernetes provider. The kinds
and omitKinds
fields on a Kubernetes account definition no longer restrict what Kubernetes kinds can be deployed by Clouddriver; instead, these fields will now only control what kinds Clouddriver caches. Spinnaker operators should ensure that Kubernetes RBAC controls are used to restrict what kinds Spinnaker can deploy.
spinnaker/front50#1252 adds optional query parameters for enabledPipelines (boolean), enabledTriggers (boolean), and triggerTypes (string) to the GET /pipelines endpoint and a new GET /pipelines/{application:.+}/name/{name:.+} endpoint to get one pipeline by application and name.
spinnaker/front50#1251 adds a new GET /pipelines/triggeredBy/{id:.+}/{status} endpoint.
spinnaker/front50#1249 adds three new configuration flags for each object type under service-storage.
-
storage-service.*.cacheHealthCheckTimeoutSeconds: The cache is considered healthy if it's been refreshed in
cacheHealthCheckTimeoutSeconds
seconds. The default is 90. -
storage-service.*.synchronizeCacheRefresh: When true, if multiple threads attempt to refresh the cache in StorageServiceSupport simultaneously, only one actually does the refresh. The others wait until it's complete. This reduces load on the data store. The default is false.
-
storage-service.*.optimizeCacheRefreshes: Only implemented for sql data stores. When true, for objects that support versioning, cache refreshes only query the data store for objects modified (or deleted) since the last refresh. The default is false.
synchronizeCacheRefresh
and optimizeCacheRefreshes
improve the performance of keeping front50's in-memory cache up to date. Here's an example of setting them to true for both applications and pipelines.
storage-service:
application:
optimizeCacheRefreshes: true
synchronizeCacheRefresh: true
pipeline:
optimizeCacheRefreshes: true
synchronizeCacheRefresh: true