# deploy the FaaS sample in the Symhony client container and copy related files (input.txt myfunc.py) to folder /share.

# assumes you have a working kubeconfig, not required if operating in-cluster
go build -o sym-client-controller .

# copy sym-client-controller sym.sh sym_monitor.sh into the /opt/ibm/sym-client-controller folder of the Symhony client container, grant 775 permission for the .sh files and then run the following command:
# to handle the requests for a namespace only, set the namespace name with environment variable like "export NAMESPACE=dept-a"
./sym-client-controller

# create a CustomResourceDefinition
kubectl create -f artifacts/examples/crd.yaml

# create a custom resource of type SessionJob to run tasks
# user can specify the task input/output and custom function as following:
#  taskInput: /shared/input.txt
#  taskOutput: /shared/output.txt
#  taskFunction: /shared/myfunc.py
#
# the content in taskInput is from user, Symphony client will create tasks based on it
# the taskOuput is used to record the output from Symphony client
# the taskFunction is the function that defined by user to handle the tasks from taskInput
#
# to create a request for a specific namespace, add the namespace name at the end of the command like "-n dept-a"
kubectl create -f artifacts/examples/example-sessionjob.yaml

# check task status through the custom resource
kubectl describe SessionJob

Part of the result is like:

Spec:
  Deployment Name:  example-sessionjob
  Replicas:         1
  Task Function:    /share/myfunc.py
  Task Input:       /share/input.txt
  Task Output:      /share/output.txt
Status:
  Done Tasks:     100
  Pending Tasks:  72
  Running Tasks:  8