Tags: nabuskey/clouddriver
Tags
fix(api): create clouddriver-api and move cloud provider API (spinnak… …er#4694) * fix(api): create clouddriver-api and move cloud provider API This moves the current cloud provider API to the new clouddriver-api module. Not everything that is API was moved, but it is almost everything I used in the Nomad plugin. Most of what was moved has been left as is for now. The things not yet moved or modified were mainly due to the need to limit external dependencies in clouddriver-api. There are specific notes below. build.gradle doesn't add any of the cross module stuff to clouddriver-api just like clouddriver-bom. Several modules now depend on clouddriver-api, but some of thier other interproject depedencies can be removed soon (especially for the cloud providers). Modifications: * DefaultCacheData was split into DefaultCacheData and DefaultJsonCacheData to separate out the jackson based construction. * AbstractAccountCredentials was separated from AccountCredentials to remove the getPermissions method (which requires fiat) * OnDemandMetricsSupportable is a new interface for OnDemandMetricsSupport which requires spectator, but only for implementation. * SearchableProvider no longer uses groovy.transform.Canonical. It shouldn't matter since it is a Java class. I've fix a few places in the code that these changes affect. DefaultJsonCacheData is used for object mappers. AccountCredentials implemenations now extend AbstractAccountCredentials. DefaultAllowedAccountsValidator has added conditional logic around getPermissions() calls because that isn't in the AccountCredentials interface. Todo: * DescriptionValidator uses org.springframework.validation.Errors from spring-context in the method contract. * DeployHandler uses DeploymentResult which is groovy and should be refactored separately. * A few unit tests need to move to clouddriver. I didn't know if we were cool with Groovy unit tests or if I needed to rewrite them. They work where they are. This was the least invasive commit I could make. The above todos will happen in follow up PRs.
fix(docs): Update broken link (spinnaker#4711) This commit fixes a broken link to the documentation on how to submit a patch.
fix(provider/cf): fix destroy cf service instance (spinnaker#4678) (s… …pinnaker#4683) Co-authored-by: Zach Smith <zachsmith@ip-10-0-0-212.us-west-2.compute.internal> (cherry picked from commit e9f7c3e) Co-authored-by: Zach Smith <33258732+zachsmith1@users.noreply.github.com>
chore(dependencies): Autobump fiatVersion (spinnaker#4708) Co-authored-by: root <root@3d8bb27683f8>
Revert "fix(awx/titus): Server group names should be lower case (spin… …naker#4660)" (spinnaker#4663) This reverts commit 219e910.
fix(provider/cf): fix stop server group state checking (spinnaker#4661)… … (spinnaker#4662) Co-authored-by: Zach Smith <zachsmith@ip-192-168-1-69.us-west-2.compute.internal> (cherry picked from commit e7152e0) Co-authored-by: Zach Smith <33258732+zachsmith1@users.noreply.github.com>
feat(aws): Only collect names of security groups in AWS deploy descri… …ption (spinnaker#4652)
fix(kubernetes/v1): Reloading namespaces during loadData since new na… …mespaces does not reflect in HA enabled clouddriver. (spinnaker#4614) Co-authored-by: Luke Abraham <luke.abraham@lowes.com> Co-authored-by: Ethan Rogers <ethanfrogers@users.noreply.github.com>
fix(provider/cf): fix buildOnDemandCacheData (spinnaker#4624) (spinna… …ker#4626) * fix(provider/cf): fix buildOnDemandCacheData CFSGcachingAgent was using the buildOnDemandCacheData method from the abstract class it extends from. The object mapper that was used in the abstract class doesn't serialize the object correctly because it only serializes fields based on the view. This creates a problem when trying to serialize a CacheData object. The fix is to override this method and use the same logic, but use the objectMapper thats currently being used in the serverGroupCachingAgent class. The additional test covers and ensures that the ondemand cache object has the servergroup attributes it should have. This fix will also ensure that on demand cache is actually processed and not just skipped over because the resulting fields are null * switch to guava immutableMap.of Co-authored-by: Zach Smith <zachsmith@Zachs-MBP-2.attlocal.net> (cherry picked from commit 06f75ea) Co-authored-by: Zach Smith <33258732+zachsmith1@users.noreply.github.com>
Changes enter a validation of the template before creating a change s… …et or updating a stack. According to AWS SDK, it will check for template syntax only (json or yml). It will throw an exception if format is incorrect. Otherwise, validation just returns a ValidateTemplateResult. (spinnaker#4602) (spinnaker#4606) Co-authored-by: Ariadna Rouco <ariadna.rouco@adevinta.com> (cherry picked from commit ed76ff4) Co-authored-by: Aria <ariadna.rouco@gmail.com> Co-authored-by: Ethan Rogers <ethanfrogers@users.noreply.github.com>
PreviousNext