retry GetPlanning a given amount of time then exit in error

move2kube/README.md

@@ -11,6 +11,7 @@ Once the transformation is over, move2kube provides a zip file containing the tr
 ### Workflow
 ![m2k.svg](https://raw.githubusercontent.com/parodos-dev/serverless-workflows/main/move2kube/m2k.svg)
 
+Note that if an error occurs during the migration planning there is no feedback given by the move2kube instance API. To overcome this, we defined a maximum amount of retries  (`move2kube_get_plan_max_retries`) to execute while getting the planning before exiting with an error. By default the value is set to 10 and it can be overridden with the environment variable `MOVE2KUBE_GET_PLAN_MAX_RETRIES`.
 ## Components
 The use case has the following components:
 1. `m2k`: the `Sonataflow` resource representing the workflow. A matching `Deployment` is created by the sonataflow operator..
@@ -62,4 +63,4 @@ backstage-backstage   backstage-backstage-backstage-system.apps.cluster-c68jb.dy
 11. Then click on `nextStep` 
 12. Click on `run` to trigger the execution
 13. Once a new transformation has started and is waiting for your input, you will receive a notification with a link to the Q&A
-14. Once you completed the Q&A, the process will continue and the output of the transformation will be saved in your git repository, you will receive a notification to inform you of the completion of the workflow.
+14. Once you completed the Q&A, the process will continue and the output of the transformation will be saved in your git repository, you will receive a notification to inform you of the completion of the workflow.

move2kube/application.properties

@@ -6,6 +6,7 @@ mp.messaging.incoming.kogito_incoming_stream.method=POST
 
 # This property is used when sending the notification while waiting for Q&A
 move2kube_url=${MOVE2KUBE_URL:http://move2kube-svc.default.svc.cluster.local:8080}
+move2kube_get_plan_max_retries=${MOVE2KUBE_GET_PLAN_MAX_RETRIES:10}
 # This property is used to send requests to the move2kube instance
 quarkus.rest-client.move2kube_yaml.url=${MOVE2KUBE_URL:http://move2kube-svc.default.svc.cluster.local:8080}
 # This property is used to send requests to the backstage notification plugin

move2kube/m2k.sw.yml

@@ -34,6 +34,9 @@ functions:
     operation: specs/move2kube.yaml#start-transformation
   - name: createNotification
     operation: notifications#createNotification
+  - name: increasePlanRetries
+    type: expression
+    operation: ".planRetries=.planRetries + 1"
 states:
   - name: StartPlanning
     type: operation
@@ -44,6 +47,11 @@ states:
             workspace-id: ".workspaceId"
             project-id: ".projectId"
             remote-source: "\"git+\" + (.repositoryURL|sub(\"http(s)://?\";\"ssh://\")) + \"@\" + .sourceBranch"
+    transition: InitGetPlanningRetry
+  - name: InitGetPlanningRetry
+    type: inject
+    data:
+      planRetries: 0
     transition: GetPlanning
   - name: GetPlanning
     type: operation
@@ -57,13 +65,22 @@ states:
           toStateData: .
         sleep:
           before: PT2S
+    transition: IncreaseGetPlanningRetry
+  - name: IncreaseGetPlanningRetry
+    type: operation
+    actions:
+      - functionRef:
+          refName: increasePlanRetries
     transition: PlanRetrievedCheck
   - name: PlanRetrievedCheck
     type: switch
     dataConditions:
       - condition: (has("plan") and .plan != "")
         transition:
           nextState: StartTransformation
+      - condition: (.planRetries >  ($SECRET.move2kube_get_plan_max_retries | tonumber))
+        transition:
+          nextState: GetPlanRetryExceededError
     defaultCondition:
       transition: GetPlanning
   - name: StartTransformation
@@ -129,6 +146,11 @@ states:
     data:
       exitMessage: '"Error while saving transformation output. If no context, it was due to timeout expiration"'
     transition: PrintExitErrorMessage
+  - name: GetPlanRetryExceededError
+    type: inject
+    data:
+      exitMessage: '"Get plan retries exceeded, an error probably occurred, check the move2kube instance logs"'
+    transition: PrintExitErrorMessage
   - name: PrintExitErrorMessage
     type: parallel
     branches:
@@ -138,7 +160,7 @@ states:
           functionRef:
             refName: systemOut
             arguments:
-              message: '${"m2k workflow: " + $WORKFLOW.instanceId + " has finalized with error. Exit message: " + .exitMessage + " -- Context: "  + .error }'
+              message: '${"m2k workflow: " + $WORKFLOW.instanceId + " has finalized with error. Exit message: " + .exitMessage + " -- Plan retries: " + (.planRetries|tostring) + " -- Context: "  + .error }'
     - name: createNotification
       actions:
         - name: createNotification
@@ -147,7 +169,7 @@ states:
             refName: createNotification
             arguments:
               title: '"Move2Kube workflow " + $WORKFLOW.instanceId + " failed"'
-              message: '"Move2Kube workflow  " + $WORKFLOW.instanceId + " on workspace " + .workspaceId + " and project " + .projectId + " failed with exit message: "+ .exitMessage + "\n\nError: "  + .error'
+              message: '"Move2Kube workflow  " + $WORKFLOW.instanceId + " on workspace " + .workspaceId + " and project " + .projectId + " failed with exit message: "+ .exitMessage + " -- Plan retries: "  + (.planRetries|tostring) + " -- Error: "  + .error'
               origin: "Move2Kube Workflow"
               topic: "Move2Kube Workflow"
     end: true