Emit hive request wait time as a metric

Merge pull request #61 from lyft/hive-request-wait-time-metric

Emit Hive request wait time as a metric

Fix for retry attempt for sub-nodes (of a dynamic node)

Use sub-node retry attempt instead of parent node's (#60)

* Use sub-node retry attempt instead of parent node

* Fix Unit Test

Migrate to go mod

https://github.com/lyft/flytepropeller/pull/38/files

Dynamic Node Retry output dir should have retry attempt in the path

v0.1.32

Dynamic sub-nodes should have retry attempt in the output dir path (#59)

Fixing caching and output dir in wf event

Fixing the output dir in workflow event sent to Admin (#58)

* test

* .

* new rm

* increment attempt after finalize, just before marking it running again

* test

* updating dependencies

Fixing output path in launchplan node

fixing output path in launchplan node (#57)

* fixing output path in launchplan node

* assert on function call

Fix invalid node lookup on abort/finalize

v0.1.29

Fix invalid node lookup on abort/finalize (#56)

Simplify usage of resource namespace and make namespace consistent everywhere

Merge pull request #54 from lyft/simpler-resource-manager-namespace

Simplify usage of resource namespace and fix tests correspondingly

Adding metrics for pod delay

Adding metrics to capture pod delay in pending state (#47)

* adding metrics to capture pod delay in pending state

* typo

* .

* fix

Fix and reenable resource-version cache

• Introduce a config-based factory method to control the workflow store type.
• Change WorkflowStore interface to return the committed version of the workflow object from etcd.
• Change resource version caching to only occur when the etcd. update operation results in a resource version change.
• Fix unit tests to simulate updating resource versions.