Update quota monitor blueprint to support project discovery

blueprints/cloud-operations/quota-monitoring/README.md

@@ -38,9 +38,10 @@ The region, location of the bundle used to deploy the function, and scheduling f
 
 The `quota_config` variable mirrors the arguments accepted by the Python program, and allows configuring several different aspects of its behaviour:
 
+- `quota_config.discover_root` organization or folder to be used to discover all underlying projects to track quotas for, in `organizations/nnnnn` or `folders/nnnnn` format
 - `quota_config.exclude` do not generate metrics for quotas matching prefixes listed here
 - `quota_config.include` only generate metrics for quotas matching prefixes listed here
-- `quota_config.projects` projects to track quotas for, defaults to the project where metrics are stored
+- `quota_config.projects` projects to track quotas for, defaults to the project where metrics are stored, if projects are automatically discovered, those in this list are appended. 
 - `quota_config.regions` regions to track quotas for, defaults to the `global` region for project-level quotas
 - `dry_run` do not write actual metrics
 - `verbose` increase logging verbosity
@@ -54,7 +55,6 @@ Clone this repository or [open it in cloud shell](https://ssh.cloud.google.com/c
 - `terraform init`
 - `terraform apply -var project_id=my-project-id`
 <!-- BEGIN TFDOC -->
-
 ## Variables
 
 | name | description | type | required | default |
@@ -64,10 +64,9 @@ Clone this repository or [open it in cloud shell](https://ssh.cloud.google.com/c
 | [bundle_path](variables.tf#L33) | Path used to write the intermediate Cloud Function code bundle. | <code>string</code> |  | <code>&#34;.&#47;bundle.zip&#34;</code> |
 | [name](variables.tf#L39) | Arbitrary string used to name created resources. | <code>string</code> |  | <code>&#34;quota-monitor&#34;</code> |
 | [project_create_config](variables.tf#L45) | Create project instead of using an existing one. | <code title="object&#40;&#123;&#10;  billing_account &#61; string&#10;  parent          &#61; optional&#40;string&#41;&#10;&#125;&#41;">object&#40;&#123;&#8230;&#125;&#41;</code> |  | <code>null</code> |
-| [quota_config](variables.tf#L59) | Cloud function configuration. | <code title="object&#40;&#123;&#10;  exclude &#61; optional&#40;list&#40;string&#41;, &#91;&#10;    &#34;a2&#34;, &#34;c2&#34;, &#34;c2d&#34;, &#34;committed&#34;, &#34;g2&#34;, &#34;interconnect&#34;, &#34;m1&#34;, &#34;m2&#34;, &#34;m3&#34;,&#10;    &#34;nvidia&#34;, &#34;preemptible&#34;&#10;  &#93;&#41;&#10;  include  &#61; optional&#40;list&#40;string&#41;&#41;&#10;  projects &#61; optional&#40;list&#40;string&#41;&#41;&#10;  regions  &#61; optional&#40;list&#40;string&#41;&#41;&#10;  dry_run  &#61; optional&#40;bool, false&#41;&#10;  verbose  &#61; optional&#40;bool, false&#41;&#10;&#125;&#41;">object&#40;&#123;&#8230;&#125;&#41;</code> |  | <code>&#123;&#125;</code> |
-| [region](variables.tf#L76) | Compute region used in the example. | <code>string</code> |  | <code>&#34;europe-west1&#34;</code> |
-| [schedule_config](variables.tf#L82) | Schedule timer configuration in crontab format. | <code>string</code> |  | <code>&#34;0 &#42; &#42; &#42; &#42;&#34;</code> |
-
+| [quota_config](variables.tf#L59) | Cloud function configuration. | <code title="object&#40;&#123;&#10;  exclude &#61; optional&#40;list&#40;string&#41;, &#91;&#10;    &#34;a2&#34;, &#34;c2&#34;, &#34;c2d&#34;, &#34;committed&#34;, &#34;g2&#34;, &#34;interconnect&#34;, &#34;m1&#34;, &#34;m2&#34;, &#34;m3&#34;,&#10;    &#34;nvidia&#34;, &#34;preemptible&#34;&#10;  &#93;&#41;&#10;  discovery_root &#61; optional&#40;string, &#34;&#34;&#41;&#10;  dry_run        &#61; optional&#40;bool, false&#41;&#10;  include        &#61; optional&#40;list&#40;string&#41;&#41;&#10;  projects       &#61; optional&#40;list&#40;string&#41;&#41;&#10;  regions        &#61; optional&#40;list&#40;string&#41;&#41;&#10;  verbose        &#61; optional&#40;bool, false&#41;&#10;&#125;&#41;">object&#40;&#123;&#8230;&#125;&#41;</code> |  | <code>&#123;&#125;</code> |
+| [region](variables.tf#L85) | Compute region used in the example. | <code>string</code> |  | <code>&#34;europe-west1&#34;</code> |
+| [schedule_config](variables.tf#L91) | Schedule timer configuration in crontab format. | <code>string</code> |  | <code>&#34;0 &#42; &#42; &#42; &#42;&#34;</code> |
 <!-- END TFDOC -->
 ## Test
 
@@ -80,5 +79,5 @@ module "test" {
     billing_account = "12345-ABCDE-12345"
   }
 }
-# tftest modules=4 resources=14
+# tftest modules=4 resources=19
 ```

blueprints/cloud-operations/quota-monitoring/main.tf

@@ -20,6 +20,8 @@ locals {
     ? [var.project_id]
     : var.quota_config.projects
   )
+  discovery_root_type = split("/", coalesce(var.quota_config["discovery_root"], "/"))[0]
+  discovery_root_id   = split("/", coalesce(var.quota_config["discovery_root"], "/"))[1]
 }
 
 module "project" {
@@ -29,8 +31,11 @@ module "project" {
   parent          = try(var.project_create_config.parent, null)
   project_create  = var.project_create_config != null
   services = [
-    "compute.googleapis.com",
-    "cloudfunctions.googleapis.com"
+    "cloudasset.googleapis.com",
+    "cloudbuild.googleapis.com",
+    "cloudfunctions.googleapis.com",
+    "cloudscheduler.googleapis.com",
+    "compute.googleapis.com"
   ]
 }
 
@@ -81,6 +86,55 @@ resource "google_cloud_scheduler_job" "default" {
   }
 }
 
+resource "google_organization_iam_member" "org_asset_viewer" {
+  count  = local.discovery_root_type == "organizations" ? 1 : 0
+  org_id = local.discovery_root_id
+  role   = "roles/cloudasset.viewer"
+  member = module.cf.service_account_iam_email
+}
+
+
+# role with the least privilege including compute.projects.get permission
+resource "google_organization_iam_member" "org_network_viewer" {
+  count  = local.discovery_root_type == "organizations" ? 1 : 0
+  org_id = local.discovery_root_id
+  role   = "roles/compute.networkViewer"
+  member = module.cf.service_account_iam_email
+}
+
+resource "google_organization_iam_member" "org_quota_viewer" {
+  count  = local.discovery_root_type == "organizations" ? 1 : 0
+  org_id = local.discovery_root_id
+  role   = "roles/servicemanagement.quotaViewer"
+  member = module.cf.service_account_iam_email
+}
+
+resource "google_folder_iam_member" "folder_asset_viewer" {
+  count  = local.discovery_root_type == "folders" ? 1 : 0
+  folder = local.discovery_root_id
+  role   = "roles/cloudasset.viewer"
+  member = module.cf.service_account_iam_email
+}
+
+# role with the least privilege including compute.projects.get permission
+resource "google_folder_iam_member" "folder_network_viewer" {
+  count  = local.discovery_root_type == "folders" ? 1 : 0
+  folder = local.discovery_root_id
+  role   = "roles/compute.networkViewer"
+  member = module.cf.service_account_iam_email
+}
+
+resource "google_folder_iam_member" "folder_quota_viewer" {
+  count  = local.discovery_root_type == "folders" ? 1 : 0
+  folder = local.discovery_root_id
+  role   = "roles/servicemanagement.quotaViewer"
+  member = module.cf.service_account_iam_email
+}
+
+
+
+
+
 resource "google_project_iam_member" "metric_writer" {
   project = module.project.project_id
   role    = "roles/monitoring.metricWriter"

blueprints/cloud-operations/quota-monitoring/src/main.py

@@ -39,6 +39,9 @@
 URL_PROJECT = 'https://compute.googleapis.com/compute/v1/projects/{}'
 URL_REGION = 'https://compute.googleapis.com/compute/v1/projects/{}/regions/{}'
 URL_TS = 'https://monitoring.googleapis.com/v3/projects/{}/timeSeries'
+URL_DISCOVERY = ('https://cloudasset.googleapis.com/v1/{}/assets?'
+                 'assetTypes=cloudresourcemanager.googleapis.com%2FProject&'
+                 'contentType=RESOURCE&pageSize=100&pageToken={}')
 
 _Quota = collections.namedtuple('_Quota',
                                 'project region tstamp metric limit usage')
@@ -80,8 +83,8 @@ def _api_format(self, name, value):
     else:
       d['valueType'] = 'INT64'
       d['points'][0]['value'] = {'int64Value': value}
-    # remove this label if cardinality gets too high
-    d['metric']['labels']['quota'] = f'{self.usage}/{self.limit}'
+    # re-enable the following line if cardinality is not a problem
+    # d['metric']['labels']['quota'] = f'{self.usage}/{self.limit}'
     return d
 
   @property
@@ -92,7 +95,7 @@ def timeseries(self):
       ratio = 0
     yield self._api_format('ratio', ratio)
     yield self._api_format('usage', self.usage)
-    # yield self._api_format('limit', self.limit)
+    yield self._api_format('limit', self.limit)
 
 
 def batched(iterable, n):
@@ -112,6 +115,23 @@ def configure_logging(verbose=True):
   warnings.filterwarnings('ignore', r'.*end user credentials.*', UserWarning)
 
 
+def discover_projects(discovery_root):
+  'Discovers projects under a folder or organization.'
+  if discovery_root.partition('/')[0] not in ('folders', 'organizations'):
+    raise SystemExit(f'Invalid discovery root {discovery_root}.')
+  next_page_token = ''
+  while True:
+    list_assets_results = fetch(
+        HTTPRequest(URL_DISCOVERY.format(discovery_root, next_page_token)))
+    if 'assets' in list_assets_results:
+      for asset in list_assets_results['assets']:
+        if (asset['resource']['data']['lifecycleState'] == 'ACTIVE'):
+          yield asset['resource']['data']['projectId']
+    next_page_token = list_assets_results.get('nextPageToken')
+    if not next_page_token:
+      break
+
+
 def fetch(request, delete=False):
   'Minimal HTTP client interface for API calls.'
   logging.debug(f'fetch {"POST" if request.data else "GET"} {request.url}')
@@ -163,9 +183,13 @@ def get_quotas(project, region='global'):
 
 @click.command()
 @click.argument('project-id', required=True)
+@click.option(
+    '--discovery-root', '-dr', required=False, help=
+    'Root node used to dynamically fetch projects, in organizations/nnn or folders/nnn format.'
+)
 @click.option(
     '--project-ids', multiple=True, help=
-    'Project ids to monitor (multiple). Defaults to monitoring project if not set.'
+    'Project ids to monitor (multiple). Defaults to monitoring project if not set, values are appended to those found under discovery-root'
 )
 @click.option('--regions', multiple=True,
               help='Regions (multiple). Defaults to "global" if not set.')
@@ -175,11 +199,13 @@ def get_quotas(project, region='global'):
               help='Exclude quotas starting with keyword (multiple).')
 @click.option('--dry-run', is_flag=True, help='Do not write metrics.')
 @click.option('--verbose', is_flag=True, help='Verbose output.')
-def main_cli(project_id=None, project_ids=None, regions=None, include=None,
-             exclude=None, dry_run=False, verbose=False):
+def main_cli(project_id=None, discovery_root=None, project_ids=None,
+             regions=None, include=None, exclude=None, dry_run=False,
+             verbose=False):
   'Fetch GCE quotas and writes them as custom metrics to Stackdriver.'
   try:
-    _main(project_id, project_ids, regions, include, exclude, dry_run, verbose)
+    _main(project_id, discovery_root, project_ids, regions, include, exclude,
+          dry_run, verbose)
   except RuntimeError as e:
     logging.exception(f'exception raised: {e.args[0]}')
 
@@ -193,14 +219,18 @@ def main(event, context):
     raise
 
 
-def _main(monitoring_project, projects=None, regions=None, include=None,
-          exclude=None, dry_run=False, verbose=False):
+def _main(monitoring_project, discovery_root=None, projects=None, regions=None,
+          include=None, exclude=None, dry_run=False, verbose=False):
   """Module entry point used by cli and cloud function wrappers."""
   configure_logging(verbose=verbose)
-  projects = projects or [monitoring_project]
+
+  # default to monitoring scope project if projects parameter is not passed, then merge the list with discovered projects, if any
   regions = regions or ['global']
   include = set(include or [])
   exclude = set(exclude or [])
+  projects = projects or [monitoring_project]
+  if (discovery_root):
+    projects = set(list(projects) + list(discover_projects(discovery_root)))
   for k in ('monitoring_project', 'projects', 'regions', 'include', 'exclude'):
     logging.debug(f'{k} {locals().get(k)}')
   timeseries = []

blueprints/cloud-operations/quota-monitoring/variables.tf

@@ -63,14 +63,23 @@ variable "quota_config" {
       "a2", "c2", "c2d", "committed", "g2", "interconnect", "m1", "m2", "m3",
       "nvidia", "preemptible"
     ])
-    include  = optional(list(string))
-    projects = optional(list(string))
-    regions  = optional(list(string))
-    dry_run  = optional(bool, false)
-    verbose  = optional(bool, false)
+    discovery_root = optional(string, "")
+    dry_run        = optional(bool, false)
+    include        = optional(list(string))
+    projects       = optional(list(string))
+    regions        = optional(list(string))
+    verbose        = optional(bool, false)
   })
   nullable = false
   default  = {}
+  validation {
+    condition = (
+      var.quota_config.discovery_root == "" ||
+      startswith(var.quota_config.discovery_root, "folders/") ||
+      startswith(var.quota_config.discovery_root, "organizations/")
+    )
+    error_message = "non-null discovery root needs to start with folders/ or organizations/"
+  }
 }
 
 variable "region" {