feat: adds cleaner utility for orphaned resources (#34)

* feat: adds cleaner utility function

Co-authored-by: Benjamin E. Coe <[email protected]>

ai-platform/snippets/test/clean.js

@@ -0,0 +1,212 @@
+// Copyright 2021 Google LLC
+//
+// Licensed under the Apache License, Version 2.0 (the "License");
+// you may not use this file except in compliance with the License.
+// You may obtain a copy of the License at
+//
+//     https://www.apache.org/licenses/LICENSE-2.0
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+/**
+ * This module contains utility functions for removing unneeded, stale, or
+ * orphaned resources from test project.
+ *
+ * Removes:
+ * - Datasets
+ * - Training pipelines
+ * - Models
+ * - Endpoints
+ * - Batch prediction jobs
+ */
+const MAXIMUM_AGE = 3600000 * 24 * 2; // 2 days in milliseconds
+const TEMP_RESOURCE_PREFIX = 'temp';
+const LOCATION = 'us-central1';
+
+// All AI Platform resources need to specify a hostname.
+const clientOptions = {
+  apiEndpoint: 'us-central1-aiplatform.googleapis.com',
+};
+
+/**
+ * Determines whether a resource should be deleted based upon its
+ * age and name.
+ * @param {string} displayName the display name of the resource
+ * @param {Timestamp} createTime when the resource is created
+ * @returns {bool}
+ */
+function checkDeletionStatus(displayName, createTime) {
+  const NOW = new Date();
+  // Check whether this resources is a temporary resource
+  if (displayName.indexOf(TEMP_RESOURCE_PREFIX) === -1) {
+    return false;
+  }
+
+  // Check how old the resource is
+  const ageOfResource = new Date(createTime.seconds * 1000);
+  if (NOW - ageOfResource < MAXIMUM_AGE) {
+    return false;
+  }
+
+  return true;
+}
+
+/**
+ * Removes all temporary datasets older than the maximum age.
+ * @param {string} projectId the project to remove datasets from
+ * @returns {Promise}
+ */
+async function cleanDatasets(projectId) {
+  const {DatasetServiceClient} = require('@google-cloud/aiplatform');
+  const datasetServiceClient = new DatasetServiceClient(clientOptions);
+
+  const [datasets] = await datasetServiceClient.listDatasets({
+    parent: `projects/${projectId}/locations/${LOCATION}`,
+  });
+
+  for (const dataset of datasets) {
+    const {displayName, createTime, name} = dataset;
+
+    if (checkDeletionStatus(displayName, createTime)) {
+      await datasetServiceClient.deleteDataset({
+        name,
+      });
+    }
+  }
+}
+
+/**
+ * Removes all temporary training pipelines older than the maximum age.
+ * @param {string} projectId the project to remove pipelines from
+ * @returns {Promise}
+ */
+async function cleanTrainingPipelines(projectId) {
+  const {PipelineServiceClient} = require('@google-cloud/aiplatform');
+  const pipelineServiceClient = new PipelineServiceClient(clientOptions);
+
+  const [pipelines] = await pipelineServiceClient.listTrainingPipelines({
+    parent: `projects/${projectId}/locations/${LOCATION}`,
+  });
+
+  for (const pipeline of pipelines) {
+    const {displayName, createTime, name} = pipeline;
+
+    if (checkDeletionStatus(displayName, createTime)) {
+      await pipelineServiceClient.deleteTrainingPipeline({
+        name,
+      });
+    }
+  }
+}
+
+/**
+ * Removes all temporary models older than the maximum age.
+ * @param {string} projectId the project to remove models from
+ * @returns {Promise}
+ */
+async function cleanModels(projectId) {
+  const {
+    ModelServiceClient,
+    EndpointServiceClient,
+  } = require('@google-cloud/aiplatform');
+  const modelServiceClient = new ModelServiceClient(clientOptions);
+
+  const [models] = await modelServiceClient.listModels({
+    parent: `projects/${projectId}/locations/${LOCATION}`,
+  });
+
+  for (const model of models) {
+    const {displayName, createTime, deployedModels, name} = model;
+
+    if (checkDeletionStatus(displayName, createTime)) {
+      // Need to check if model is deployed to an endpoint
+      // Undeploy the model everywhere it is deployed
+      for (const deployedModel of deployedModels) {
+        const {endpoint, deployedModelId} = deployedModel;
+
+        const endpointServiceClient = new EndpointServiceClient(clientOptions);
+        await endpointServiceClient.undeployModel({
+          endpoint,
+          deployedModelId,
+        });
+      }
+
+      await modelServiceClient.deleteModel({
+        name,
+      });
+    }
+  }
+}
+
+/**
+ * Removes all temporary endpoints older than the maximum age.
+ * @param {string} projectId the project to remove endpoints from
+ * @returns {Promise}
+ */
+async function cleanEndpoints(projectId) {
+  const {EndpointServiceClient} = require('@google-cloud/aiplatform');
+  const endpointServiceClient = new EndpointServiceClient(clientOptions);
+
+  const [endpoints] = await endpointServiceClient.listEndpoints({
+    parent: `projects/${projectId}/locations/${LOCATION}`,
+  });
+
+  for (const endpoint of endpoints) {
+    const {displayName, createTime, name} = endpoint;
+
+    if (checkDeletionStatus(displayName, createTime)) {
+      await endpointServiceClient.deleteEndpoint({
+        name,
+      });
+    }
+  }
+}
+
+/**
+ * Removes all temporary batch prediction jobs
+ * @param {string} projectId the project to remove prediction jobs from
+ * @returns {Promise}
+ */
+async function cleanBatchPredictionJobs(projectId) {
+  const {JobServiceClient} = require('@google-cloud/aiplatform');
+  const jobServiceClient = new JobServiceClient(clientOptions);
+
+  const [batchPredictionJobs] = await jobServiceClient.listBatchPredictionJobs({
+    parent: `projects/${projectId}/locations/${LOCATION}`,
+  });
+
+  for (const job of batchPredictionJobs) {
+    const {displayName, createTime, name} = job;
+    if (checkDeletionStatus(displayName, createTime)) {
+      await jobServiceClient.deleteBatchPredictionJob({
+        name,
+      });
+    }
+  }
+}
+
+/**
+ * Removes all of the temporary resources older than the maximum age.
+ * @param {string} projectId the project to remove resources from
+ * @returns {Promise}
+ */
+async function cleanAll(projectId) {
+  await cleanDatasets(projectId);
+  await cleanTrainingPipelines(projectId);
+  await cleanModels(projectId);
+  await cleanEndpoints(projectId);
+  await cleanBatchPredictionJobs(projectId);
+}
+
+module.exports = {
+  cleanAll: cleanAll,
+  cleanDatasets: cleanDatasets,
+  cleanTrainingPipelines: cleanTrainingPipelines,
+  cleanModels: cleanModels,
+  cleanEndpoints: cleanEndpoints,
+  cleanBatchPredictionJobs: cleanBatchPredictionJobs,
+};

ai-platform/snippets/test/create-training-pipeline-image-classification.test.js

@@ -17,7 +17,8 @@
 'use strict';
 
 const {assert} = require('chai');
-const {after, describe, it} = require('mocha');
+const {after, before, describe, it} = require('mocha');
+const clean = require('./clean');
 
 const uuid = require('uuid').v4;
 const cp = require('child_process');
@@ -41,6 +42,10 @@ const location = process.env.LOCATION;
 let trainingPipelineId;
 
 describe('AI platform create training pipeline image classification', () => {
+  before('should delete any old and/or orphaned resources', async () => {
+    await clean.cleanTrainingPipelines(project);
+  });
+
   it('should create a new image classification training pipeline', async () => {
     const stdout = execSync(
       `node ./create-training-pipeline-image-classification.js ${datasetId} ${modelDisplayName} ${trainingPipelineDisplayName} ${project} ${location}`