Merge pull request #32 from maazmaqsood/dynamic-base-image

Enabling Dynamic Addition of Images for Dockerization

.dockerignore

@@ -3,5 +3,7 @@
 !examples/multi_backends/*.py
 !examples/wiki_parser/*.py
 !examples/caching/*.py
+!examples/etl_pipeline/*.py
+!examples/utilizing_gpu/*.py
 !bin
 !pirlib

examples/etl_pipeline/argo-train.yml

@@ -0,0 +1,105 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Workflow
+metadata:
+  generateName: argo-train-
+spec:
+  entrypoint: DAG-etl-pipeline
+  templates:
+  - name: create-spark-session-template
+    container:
+      image: maaz112233/pirlib
+      command:
+      - python
+      - -m
+      - pirlib.backends.argo_batch
+      - node
+      - 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
+      - gASVigAAAAAAAABdlIwKcGlybGliLnBpcpSMCkdyYXBoSW5wdXSUk5QpgZR9lCiMAmlklIwHZGF0YXNldJSMBmlvdHlwZZSMCURJUkVDVE9SWZSMBG1ldGGUaAGMCE1ldGFkYXRhlJOUKYGUfZQojARuYW1llIwHZGF0YXNldJSMC2Fubm90YXRpb25zlE51YnViYS4=
+      volumeMounts:
+      - name: node-outputs
+        mountPath: /mnt/node_outputs
+      - name: dataset
+        mountPath: /mnt/graph_inputs/dataset
+    volumes:
+    - name: node-outputs
+      nfs:
+        server: k8s-master.cm.cluster
+        path: /home/maaz/pirlib/examples/etl_pipeline/outputs
+        readOnly: no
+    - name: dataset
+      nfs:
+        server: k8s-master.cm.cluster
+        path: /home/maaz/pirlib/examples/etl_pipeline/inputs
+        readOnly: yes
+  - name: extract-transform-load-template
+    container:
+      image: maaz112233/pirlib
+      command:
+      - python
+      - -m
+      - pirlib.backends.argo_batch
+      - node
+      - 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
+      - gASVigAAAAAAAABdlIwKcGlybGliLnBpcpSMCkdyYXBoSW5wdXSUk5QpgZR9lCiMAmlklIwHZGF0YXNldJSMBmlvdHlwZZSMCURJUkVDVE9SWZSMBG1ldGGUaAGMCE1ldGFkYXRhlJOUKYGUfZQojARuYW1llIwHZGF0YXNldJSMC2Fubm90YXRpb25zlE51YnViYS4=
+      volumeMounts:
+      - name: node-outputs
+        mountPath: /mnt/node_outputs
+      - name: dataset
+        mountPath: /mnt/graph_inputs/dataset
+    volumes:
+    - name: node-outputs
+      nfs:
+        server: k8s-master.cm.cluster
+        path: /home/maaz/pirlib/examples/etl_pipeline/outputs
+        readOnly: no
+    - name: dataset
+      nfs:
+        server: k8s-master.cm.cluster
+        path: /home/maaz/pirlib/examples/etl_pipeline/inputs
+        readOnly: yes
+  - name: etl-pipeline-template
+    container:
+      image: maaz112233/pirlib
+      command:
+      - python
+      - -m
+      - pirlib.backends.argo_batch
+      - graph
+      - gASVBAEAAAAAAABdlIwKcGlybGliLnBpcpSMC0dyYXBoT3V0cHV0lJOUKYGUfZQojAJpZJSMBnJldHVybpSMBmlvdHlwZZSMCURJUkVDVE9SWZSMBnNvdXJjZZRoAYwKRGF0YVNvdXJjZZSTlCmBlH2UKIwHbm9kZV9pZJSMFmV4dHJhY3RfdHJhbnNmb3JtX2xvYWSUjAtzdWJncmFwaF9pZJROjAlvdXRwdXRfaWSUjAZyZXR1cm6UjA5ncmFwaF9pbnB1dF9pZJROdWKMBG1ldGGUaAGMCE1ldGFkYXRhlJOUKYGUfZQojARuYW1llIwGcmV0dXJulIwLYW5ub3RhdGlvbnOUTnVidWJhLg==
+      volumeMounts:
+      - name: node-outputs
+        mountPath: /mnt/node_outputs
+      - name: dataset
+        mountPath: /mnt/graph_inputs/dataset
+      - name: graph-outputs
+        mountPath: /mnt/graph_outputs
+    volumes:
+    - name: node-outputs
+      nfs:
+        server: k8s-master.cm.cluster
+        path: /home/maaz/pirlib/examples/etl_pipeline/outputs
+        readOnly: no
+    - name: dataset
+      nfs:
+        server: k8s-master.cm.cluster
+        path: /home/maaz/pirlib/examples/etl_pipeline/inputs
+        readOnly: yes
+    - name: graph-outputs
+      nfs:
+        server: k8s-master.cm.cluster
+        path: /home/maaz/pirlib/examples/etl_pipeline/outputs
+        readOnly: no
+  - name: DAG-etl-pipeline
+    dag:
+      tasks:
+      - name: create-spark-session
+        template: create-spark-session-template
+        dependencies: []
+      - name: extract-transform-load
+        template: extract-transform-load-template
+        dependencies: []
+      - name: etl-pipeline
+        template: etl-pipeline-template
+        dependencies:
+        - create-spark-session
+        - extract-transform-load

examples/etl_pipeline/etl.py

@@ -0,0 +1,60 @@
+from dataclasses import asdict
+import requests
+import pandas as pd
+from pirlib.iotypes import DirectoryPath, FilePath
+from pirlib.pipeline import pipeline
+from pirlib.task import task
+from pyspark.sql import SparkSession
+
+"""
+Python Extract Transform Load Pipeline Example
+"""
+
+
+@task
+def create_spark_session(dataset: DirectoryPath) -> DirectoryPath:
+    # Create a Spark session
+    spark = SparkSession.builder.appName("Simple Spark Session").getOrCreate()
+    # Check if the Spark session is successfully created
+    print("Spark version:", spark.version)
+    # Perform some simple operations using the Spark session
+    data = [("Alice", 34), ("Bob", 45), ("Charlie", 29)]
+    df = spark.createDataFrame(data, ["Name", "Age"])
+    # Show the DataFrame
+    df.show()
+    # Stop the Spark session when done
+    spark.stop()
+    return dataset
+
+
+@task
+def extract_transform_load(dataset: DirectoryPath) -> DirectoryPath:
+    """This API extracts data from
+    http://universities.hipolabs.com
+    """
+    API_URL = "http://universities.hipolabs.com/search?country=United+States"
+    data = requests.get(API_URL).json()
+
+    df = pd.DataFrame(data)
+    print(f"Total Number of universities from API {len(data)}")
+
+    df = df[df["name"].str.contains("California")]
+    print(f"Number of universities in california {len(df)}")
+
+    df["domains"] = [",".join(map(str, l)) for l in df["domains"]]
+    df["web_pages"] = [",".join(map(str, l)) for l in df["web_pages"]]
+    df = df.reset_index(drop=True)
+    df = df[["domains", "country", "web_pages", "name"]]
+
+    outdir = task.context().output
+    file_name = outdir / "file.csv"
+    df.to_csv(file_name, sep="\t", encoding="utf-8")
+
+    return outdir
+
+
+@pipeline
+def etl_pipeline(dataset: DirectoryPath) -> DirectoryPath:
+    create_spark_session(dataset)
+    data = extract_transform_load(dataset)
+    return data

examples/etl_pipeline/package_argo.yml

@@ -0,0 +1,86 @@
+graphs:
+- id: etl_pipeline
+  nodes:
+  - id: create_spark_session
+    entrypoints:
+      main:
+        version: v1
+        handler: examples.etl_pipeline.etl:create_spark_session
+        runtime: python:3.8
+        codeurl: null
+        image: maaz112233/pirlib
+    framework: null
+    config:
+      timer: false
+    inputs:
+    - id: dataset
+      iotype: DIRECTORY
+      source:
+        node_id: null
+        subgraph_id: null
+        output_id: null
+        graph_input_id: dataset
+      meta:
+        name: dataset
+        annotations: null
+    outputs:
+    - id: return
+      iotype: DIRECTORY
+      meta:
+        name: return
+        annotations: null
+    meta:
+      name: create_spark_session
+      annotations: null
+  - id: extract_transform_load
+    entrypoints:
+      main:
+        version: v1
+        handler: examples.etl_pipeline.etl:extract_transform_load
+        runtime: python:3.8
+        codeurl: null
+        image: maaz112233/pirlib
+    framework: null
+    config:
+      timer: false
+    inputs:
+    - id: dataset
+      iotype: DIRECTORY
+      source:
+        node_id: null
+        subgraph_id: null
+        output_id: null
+        graph_input_id: dataset
+      meta:
+        name: dataset
+        annotations: null
+    outputs:
+    - id: return
+      iotype: DIRECTORY
+      meta:
+        name: return
+        annotations: null
+    meta:
+      name: extract_transform_load
+      annotations: null
+  subgraphs: []
+  inputs:
+  - id: dataset
+    iotype: DIRECTORY
+    meta:
+      name: dataset
+      annotations: null
+  outputs:
+  - id: return
+    iotype: DIRECTORY
+    source:
+      node_id: extract_transform_load
+      subgraph_id: null
+      output_id: return
+      graph_input_id: null
+    meta:
+      name: return
+      annotations: null
+  meta:
+    name: etl_pipeline
+    annotations: null

examples/etl_pipeline/requirements.txt

@@ -0,0 +1,2 @@
+requests
+pandas

examples/etl_pipeline/run_argo.sh

@@ -0,0 +1,26 @@
+EXAMPLEDIR=$(dirname $0)
+ROOTDIR=$EXAMPLEDIR/../..
+
+### Module 1: Docker_Packaging
+python $ROOTDIR/bin/pircli dockerize \
+    $ROOTDIR \
+	--auto \
+	--pipeline examples.etl_pipeline.etl:etl_pipeline \
+	--output $EXAMPLEDIR/package_argo.yml \
+	--flatten \
+	--docker_base_image godatadriven/pyspark:latest
+
+# Convert EXAMPLEDIR to absolute path since docker can't bind-mount relative paths.
+EXAMPLEDIR=$([[ $EXAMPLEDIR = /* ]] && echo "$EXAMPLEDIR" || echo "$PWD/${EXAMPLEDIR#./}")
+
+### Module 2: Argoize_Module
+INPUT_dataset=$EXAMPLEDIR/inputs \
+OUTPUT=$EXAMPLEDIR/outputs \
+NFS_SERVER=k8s-master.cm.cluster \
+python  $ROOTDIR/bin/pircli generate $EXAMPLEDIR/package_argo.yml \
+	--target pirlib.backends.argo_batch:ArgoBatchBackend \
+	--output $EXAMPLEDIR/argo-train.yml
+
+# Run the Argo workflow
+argo submit -n argo --watch $EXAMPLEDIR/argo-train.yml
+