Merge ~chad.smith/cloud-init:feature/command-cloud-id into cloud-init:master

[Chad Smith]

tools: Add cloud-id command line utility

Add a quick cloud lookup utility in order to more easily determine
the cloud on which an instance is running.

The utility parses standardized attributes from
/run/cloud-init/instance-data.json to print the canonical cloud-id
for the instance. It uses known region maps if necessary to determine
on which specific cloud the instance is running.

Examples:
aws, aws-gov, aws-china, rackspace, azure-china, lxd, openstack, unknown

[Server Team CI bot (server-team-bot)]

FAILED: Continuous integration, rev:abc12004af1c150d5f3504766ed858373ed254d5
https://jenkins.ubuntu.com/server/job/cloud-init-ci/381/
Executed test runs:
    SUCCESS: Checkout
    FAILED: Unit & Style Tests

Click here to trigger a rebuild:
https://jenkins.ubuntu.com/server/job/cloud-init-ci/381/rebuild

review: Needs Fixing (continuous-integration)

Unmerged commits

abc1200... by Chad Smith on 2018-10-09

return platform from canonical_cloud_id if cloud_name is unknown

0af64de... by Chad Smith on 2018-10-09

canonical_cloud_id handles None and unknown regions

ae16154... by Chad Smith on 2018-10-09

move cloud_id out of tools and under cloudinit.cmd

8bc2628... by Chad Smith on 2018-10-09

revert tox changes

c3a9e4c... by Chad Smith on 2018-10-09

oops

655bb65... by Chad Smith on 2018-10-09

update docs, wrap timer around ec2 crawl_metadata/get_data

4981999... by Chad Smith on 2018-10-09

add tools dir to test tox envs for testing

a1d0620... by Chad Smith on 2018-10-09

tests: add test_cloud_id. fix setup.py to source cloud_id

5fef4ae... by Chad Smith on 2018-10-08

tools: add cloud-id utility to report canonical cloud info

94e3f12... by Chad Smith on 2018-10-08

tests: nocloud_kvm integration check for config-disk subplatform

diff --git a/cloudinit/cmd/cloud_id.py b/cloudinit/cmd/cloud_id.py
new file mode 100755
index 0000000..8312d08
--- /dev/null
+++ b/cloudinit/cmd/cloud_id.py
@@ -0,0 +1,88 @@
+# This file is part of cloud-init. See LICENSE file for license information.
+
+"""Commandline utility to list the canonical cloud-id for an instance."""
+
+import argparse
+import os
+import six
+import sys
+
+from cloudinit import util
+from cloudinit.sources import (
+    INSTANCE_JSON_FILE, METADATA_UNKNOWN, canonical_cloud_id)
+
+DEFAULT_INSTANCE_JSON = '/run/cloud-init/%s' % INSTANCE_JSON_FILE
+
+NAME = 'cloud-id'
+
+
+def get_parser(parser=None):
+    """Build or extend an arg parser for the cloud-id utility.
+
+    @param parser: Optional existing ArgumentParser instance representing the
+        query subcommand which will be extended to support the args of
+        this utility.
+
+    @returns: ArgumentParser with proper argument configuration.
+    """
+    if not parser:
+        parser = argparse.ArgumentParser(
+            prog=NAME,
+            description='Report the canonical cloud-id for this instance')
+    parser.add_argument(
+        '-j', '--json', action='store_true', default=False,
+        help='Report all standardized cloud-id information as json.')
+    parser.add_argument(
+        '-l', '--long', action='store_true', default=False,
+        help='Report extended cloud-id information as hyphenated string.')
+    parser.add_argument(
+        '-i', '--instance-data', type=str, default=DEFAULT_INSTANCE_JSON,
+        help=('Path to instance-data.json file. Default is %s' %
+              DEFAULT_INSTANCE_JSON))
+    return parser
+
+def error(msg):
+   sys.stderr.write('ERROR: %s\n' % msg)
+   return 1
+
+
+def handle_args(name, args):
+    """Handle calls to 'cloud-id' cli.
+
+    Print the canonical cloud-id on which the instance is running.
+
+    @return: 0 on success, 1 otherwise.
+    """
+    try:
+        instance_json = util.load_file(args.instance_data)
+    except IOError:
+        return error(
+            "File not found '%s'. Provide a path to instance data json file"
+            ' using --instance-data' % args.instance_data)
+    instance_data = util.load_json(instance_json)
+    v1 = instance_data.get('v1', {})
+    cloud_id = canonical_cloud_id(
+        v1.get('cloud_name', METADATA_UNKNOWN),
+        v1.get('region', METADATA_UNKNOWN),
+        v1.get('platform', METADATA_UNKNOWN))
+    if args.json:
+        v1['cloud_id'] = cloud_id
+        response = util.json_dumps(v1)
+    elif args.long:
+        response = '%s-%s' % (cloud_id, v1.get('region', METADATA_UNKNOWN))
+    else:
+        response = cloud_id
+    sys.stdout.write('%s\n' % response)
+    return 0
+
+
+def main():
+    """Tool to query specific instance-data values."""
+    parser = get_parser()
+    sys.exit(handle_args(NAME, parser.parse_args()))
+
+
+if __name__ == '__main__':
+    main()
+
+# vi: ts=4 expandtab
diff --git a/cloudinit/cmd/tests/test_cloud_id.py b/cloudinit/cmd/tests/test_cloud_id.py
new file mode 100644
index 0000000..a42cfa8
--- /dev/null
+++ b/cloudinit/cmd/tests/test_cloud_id.py
@@ -0,0 +1,112 @@
+# This file is part of cloud-init. See LICENSE file for license information.
+
+"""Tests for cloud-id command line utility."""
+
+from cloudinit import util
+from collections import namedtuple
+from six import StringIO
+
+from cloudinit.cmd import cloud_id
+
+from cloudinit.tests.helpers import CiTestCase, mock
+
+
+class TestCloudId(CiTestCase):
+
+    args = namedtuple('cloudidargs', ('instance_data json long'))
+
+    def setUp(self):
+        super(TestCloudId, self).setUp()
+        self.tmp = self.tmp_dir()
+        self.instance_data = self.tmp_path('instance-data.json', dir=self.tmp)
+
+    def test_cloud_id_arg_parser_defaults(self):
+        """Validate the argument defaults when not provided by the end-user."""
+        cmd = ['cloud-id']
+        with mock.patch('sys.argv', cmd):
+            args = cloud_id.get_parser().parse_args()
+        self.assertEqual('/run/cloud-init/instance-data.json', args.instance_data)
+        self.assertEqual(False, args.long)
+        self.assertEqual(False, args.json)
+
+    def test_cloud_id_arg_parse_overrides(self):
+        """Override argument defaults by specifying values for each param."""
+        util.write_file(self.instance_data, '{}')
+        cmd = ['cloud-id', '--instance-data', self.instance_data, '--long',
+               '--json']
+        with mock.patch('sys.argv', cmd):
+            args = cloud_id.get_parser().parse_args()
+        self.assertEqual(self.instance_data, args.instance_data)
+        self.assertEqual(True, args.long)
+        self.assertEqual(True, args.json)
+
+    def test_cloud_id_missing_instance_data_json(self):
+        """Exit error when the provided instance-data.json does not exist."""
+        cmd = ['cloud-id', '--instance-data', self.instance_data]
+        with mock.patch('sys.argv', cmd):
+            with mock.patch('sys.stderr', new_callable=StringIO) as m_stderr:
+                with self.assertRaises(SystemExit) as context_manager:
+                    cloud_id.main()
+        self.assertEqual(1, context_manager.exception.code)
+        self.assertIn(
+            "ERROR: File not found '%s'" % self.instance_data,
+            m_stderr.getvalue())
+
+    def test_cloud_id_from_cloud_name_in_instance_data(self):
+        """Report canonical cloud-id from cloud_name in instance-data."""
+        util.write_file(
+            self.instance_data,
+            '{"v1": {"cloud_name": "mycloud", "region": "somereg"}}')
+        cmd = ['cloud-id', '--instance-data', self.instance_data]
+        with mock.patch('sys.argv', cmd):
+            with mock.patch('sys.stdout', new_callable=StringIO) as m_stdout:
+                with self.assertRaises(SystemExit) as context_manager:
+                    cloud_id.main()
+        self.assertEqual(0, context_manager.exception.code)
+        self.assertEqual("mycloud\n", m_stdout.getvalue())
+
+    def test_cloud_id_long_name_from_instance_data(self):
+        """Report long cloud-id format from cloud_name and region."""
+        util.write_file(
+            self.instance_data,
+            '{"v1": {"cloud_name": "mycloud", "region": "somereg"}}')
+        cmd = ['cloud-id', '--instance-data', self.instance_data, '--long']
+        with mock.patch('sys.argv', cmd):
+            with mock.patch('sys.stdout', new_callable=StringIO) as m_stdout:
+                with self.assertRaises(SystemExit) as context_manager:
+                    cloud_id.main()
+        self.assertEqual(0, context_manager.exception.code)
+        self.assertEqual("mycloud-somereg\n", m_stdout.getvalue())
+
+    def test_cloud_id_lookup_from_instance_data_region(self):
+        """Report discovered canonical cloud_id when region lookup matches."""
+        util.write_file(
+            self.instance_data,
+            '{"v1": {"cloud_name": "aws", "region": "cn-north-1",'
+            ' "platform": "ec2"}}')
+        cmd = ['cloud-id', '--instance-data', self.instance_data, '--long']
+        with mock.patch('sys.argv', cmd):
+            with mock.patch('sys.stdout', new_callable=StringIO) as m_stdout:
+                with self.assertRaises(SystemExit) as context_manager:
+                    cloud_id.main()
+        self.assertEqual(0, context_manager.exception.code)
+        self.assertEqual("aws-china-cn-north-1\n", m_stdout.getvalue())
+
+    def test_cloud_id_lookup_json_instance_data_adds_cloud_id_to_json(self):
+        """Report v1 instance-data content with cloud_id when --json set."""
+        util.write_file(
+            self.instance_data,
+            '{"v1": {"cloud_name": "unknown", "region": "dfw",'
+            ' "platform": "openstack", "public_ssh_keys": []}}')
+        expected = util.json_dumps({
+            'cloud_id': 'rackspace', 'cloud_name': 'unknown',
+            'platform': 'openstack', 'public_ssh_keys': [], 'region': 'dfw'})
+        cmd = ['cloud-id', '--instance-data', self.instance_data, '--json']
+        with mock.patch('sys.argv', cmd):
+            with mock.patch('sys.stdout', new_callable=StringIO) as m_stdout:
+                with self.assertRaises(SystemExit) as context_manager:
+                    cloud_id.main()
+        self.assertEqual(0, context_manager.exception.code)
+        self.assertEqual(expected + '\n', m_stdout.getvalue())
+
+# vi: ts=4 expandtab
diff --git a/cloudinit/sources/DataSourceAliYun.py b/cloudinit/sources/DataSourceAliYun.py
index 858e082..45cc9f0 100644
--- a/cloudinit/sources/DataSourceAliYun.py
+++ b/cloudinit/sources/DataSourceAliYun.py
@@ -1,7 +1,5 @@
 # This file is part of cloud-init. See LICENSE file for license information.
 
-import os
-
 from cloudinit import sources
 from cloudinit.sources import DataSourceEc2 as EC2
 from cloudinit import util
@@ -18,25 +16,17 @@ class DataSourceAliYun(EC2.DataSourceEc2):
     min_metadata_version = '2016-01-01'
     extended_metadata_versions = []
 
-    def __init__(self, sys_cfg, distro, paths):
-        super(DataSourceAliYun, self).__init__(sys_cfg, distro, paths)
-        self.seed_dir = os.path.join(paths.seed_dir, "AliYun")
-
     def get_hostname(self, fqdn=False, resolve_ip=False, metadata_only=False):
         return self.metadata.get('hostname', 'localhost.localdomain')
 
     def get_public_ssh_keys(self):
         return parse_public_keys(self.metadata.get('public-keys', {}))
 
-    @property
-    def cloud_platform(self):
-        if self._cloud_platform is None:
-            if _is_aliyun():
-                self._cloud_platform = EC2.Platforms.ALIYUN
-            else:
-                self._cloud_platform = EC2.Platforms.NO_EC2_METADATA
-
-        return self._cloud_platform
+    def _get_cloud_name(self):
+        if _is_aliyun():
+            return EC2.CloudNames.ALIYUN
+        else:
+            return EC2.CloudNames.NO_EC2_METADATA
 
 
 def _is_aliyun():
diff --git a/cloudinit/sources/DataSourceAltCloud.py b/cloudinit/sources/DataSourceAltCloud.py
index 8cd312d..5270fda 100644
--- a/cloudinit/sources/DataSourceAltCloud.py
+++ b/cloudinit/sources/DataSourceAltCloud.py
@@ -89,7 +89,9 @@ class DataSourceAltCloud(sources.DataSource):
         '''
         Description:
             Get the type for the cloud back end this instance is running on
-            by examining the string returned by reading the dmi data.
+            by examining the string returned by reading either:
+                CLOUD_INFO_FILE or
+                the dmi data.
 
         Input:
             None
@@ -99,7 +101,14 @@ class DataSourceAltCloud(sources.DataSource):
             'RHEV', 'VSPHERE' or 'UNKNOWN'
 
         '''
-
+        if os.path.exists(CLOUD_INFO_FILE):
+            try:
+                cloud_type = util.load_file(CLOUD_INFO_FILE).strip().upper()
+            except IOError:
+                util.logexc(LOG, 'Unable to access cloud info file at %s.',
+                            CLOUD_INFO_FILE)
+                return 'UNKNOWN'
+            return cloud_type
         system_name = util.read_dmi_data("system-product-name")
         if not system_name:
             return 'UNKNOWN'
@@ -134,15 +143,7 @@ class DataSourceAltCloud(sources.DataSource):
 
         LOG.debug('Invoked get_data()')
 
-        if os.path.exists(CLOUD_INFO_FILE):
-            try:
-                cloud_type = util.load_file(CLOUD_INFO_FILE).strip().upper()
-            except IOError:
-                util.logexc(LOG, 'Unable to access cloud info file at %s.',
-                            CLOUD_INFO_FILE)
-                return False
-        else:
-            cloud_type = self.get_cloud_type()
+        cloud_type = self.get_cloud_type()
 
         LOG.debug('cloud_type: %s', str(cloud_type))
 
@@ -161,6 +162,15 @@ class DataSourceAltCloud(sources.DataSource):
         util.logexc(LOG, 'Failed accessing user data.')
         return False
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata details."""
+        cloud_type = self.get_cloud_type()
+        if not hasattr(self, 'source'):
+            self.source = sources.METADATA_UNKNOWN
+        if cloud_type == 'RHEV':
+            self.source = '/dev/fd0'
+        return '%s (%s)' % (cloud_type.lower(), self.source)
+
     def user_data_rhevm(self):
         '''
         RHEVM specific userdata read
@@ -232,6 +242,7 @@ class DataSourceAltCloud(sources.DataSource):
             try:
                 return_str = util.mount_cb(cdrom_dev, read_user_data_callback)
                 if return_str:
+                    self.source = cdrom_dev
                     break
             except OSError as err:
                 if err.errno != errno.ENOENT:
diff --git a/cloudinit/sources/DataSourceAzure.py b/cloudinit/sources/DataSourceAzure.py
index 783445e..39391d0 100644
--- a/cloudinit/sources/DataSourceAzure.py
+++ b/cloudinit/sources/DataSourceAzure.py
@@ -351,6 +351,14 @@ class DataSourceAzure(sources.DataSource):
         metadata['public-keys'] = key_value or pubkeys_from_crt_files(fp_files)
         return metadata
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        if self.seed.startswith('/dev'):
+            subplatform_type = 'config-disk'
+        else:
+            subplatform_type = 'seed-dir'
+        return '%s (%s)' % (subplatform_type, self.seed)
+
     def crawl_metadata(self):
         """Walk all instance metadata sources returning a dict on success.
 
diff --git a/cloudinit/sources/DataSourceBigstep.py b/cloudinit/sources/DataSourceBigstep.py
index 699a85b..52fff20 100644
--- a/cloudinit/sources/DataSourceBigstep.py
+++ b/cloudinit/sources/DataSourceBigstep.py
@@ -36,6 +36,10 @@ class DataSourceBigstep(sources.DataSource):
         self.userdata_raw = decoded["userdata_raw"]
         return True
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        return 'metadata (%s)' % get_url_from_file()
+
 
 def get_url_from_file():
     try:
diff --git a/cloudinit/sources/DataSourceCloudSigma.py b/cloudinit/sources/DataSourceCloudSigma.py
index c816f34..2955d3f 100644
--- a/cloudinit/sources/DataSourceCloudSigma.py
+++ b/cloudinit/sources/DataSourceCloudSigma.py
@@ -7,7 +7,7 @@
 from base64 import b64decode
 import re
 
-from cloudinit.cs_utils import Cepko
+from cloudinit.cs_utils import Cepko, SERIAL_PORT
 
 from cloudinit import log as logging
 from cloudinit import sources
@@ -84,6 +84,10 @@ class DataSourceCloudSigma(sources.DataSource):
 
         return True
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        return 'cepko (%s)' % SERIAL_PORT
+
     def get_hostname(self, fqdn=False, resolve_ip=False, metadata_only=False):
         """
         Cleans up and uses the server's name if the latter is set. Otherwise
diff --git a/cloudinit/sources/DataSourceConfigDrive.py b/cloudinit/sources/DataSourceConfigDrive.py
index 664dc4b..564e3eb 100644
--- a/cloudinit/sources/DataSourceConfigDrive.py
+++ b/cloudinit/sources/DataSourceConfigDrive.py
@@ -160,6 +160,18 @@ class DataSourceConfigDrive(openstack.SourceMixin, sources.DataSource):
                 LOG.debug("no network configuration available")
         return self._network_config
 
+    @property
+    def platform(self):
+        return 'openstack'
+
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        if self.seed_dir in self.source:
+            subplatform_type = 'seed-dir'
+        elif self.source.startswith('/dev'):
+            subplatform_type = 'config-disk'
+        return '%s (%s)' % (subplatform_type, self.source)
+
 
 def read_config_drive(source_dir):
     reader = openstack.ConfigDriveReader(source_dir)
diff --git a/cloudinit/sources/DataSourceEc2.py b/cloudinit/sources/DataSourceEc2.py
index 968ab3f..9ccf2cd 100644
--- a/cloudinit/sources/DataSourceEc2.py
+++ b/cloudinit/sources/DataSourceEc2.py
@@ -28,18 +28,16 @@ STRICT_ID_PATH = ("datasource", "Ec2", "strict_id")
 STRICT_ID_DEFAULT = "warn"
 
 
-class Platforms(object):
-    # TODO Rename and move to cloudinit.cloud.CloudNames
-    ALIYUN = "AliYun"
-    AWS = "AWS"
-    BRIGHTBOX = "Brightbox"
-    SEEDED = "Seeded"
+class CloudNames(object):
+    ALIYUN = "aliyun"
+    AWS = "aws"
+    BRIGHTBOX = "brightbox"
     # UNKNOWN indicates no positive id.  If strict_id is 'warn' or 'false',
     # then an attempt at the Ec2 Metadata service will be made.
-    UNKNOWN = "Unknown"
+    UNKNOWN = "unknown"
     # NO_EC2_METADATA indicates this platform does not have a Ec2 metadata
     # service available. No attempt at the Ec2 Metadata service will be made.
-    NO_EC2_METADATA = "No-EC2-Metadata"
+    NO_EC2_METADATA = "no-ec2-metadata"
 
 
 class DataSourceEc2(sources.DataSource):
@@ -61,8 +59,6 @@ class DataSourceEc2(sources.DataSource):
     url_max_wait = 120
     url_timeout = 50
 
-    _cloud_platform = None
-
     _network_config = sources.UNSET  # Used to cache calculated network cfg v1
 
     # Whether we want to get network configuration from the metadata service.
@@ -71,30 +67,21 @@ class DataSourceEc2(sources.DataSource):
     def __init__(self, sys_cfg, distro, paths):
         super(DataSourceEc2, self).__init__(sys_cfg, distro, paths)
         self.metadata_address = None
-        self.seed_dir = os.path.join(paths.seed_dir, "ec2")
 
     def _get_cloud_name(self):
         """Return the cloud name as identified during _get_data."""
-        return self.cloud_platform
+        return identify_platform()
 
     def _get_data(self):
-        seed_ret = {}
-        if util.read_optional_seed(seed_ret, base=(self.seed_dir + "/")):
-            self.userdata_raw = seed_ret['user-data']
-            self.metadata = seed_ret['meta-data']
-            LOG.debug("Using seeded ec2 data from %s", self.seed_dir)
-            self._cloud_platform = Platforms.SEEDED
-            return True
-
         strict_mode, _sleep = read_strict_mode(
             util.get_cfg_by_path(self.sys_cfg, STRICT_ID_PATH,
                                  STRICT_ID_DEFAULT), ("warn", None))
 
-        LOG.debug("strict_mode: %s, cloud_platform=%s",
-                  strict_mode, self.cloud_platform)
-        if strict_mode == "true" and self.cloud_platform == Platforms.UNKNOWN:
+        LOG.debug("strict_mode: %s, cloud_name=%s cloud_platform=%s",
+                  strict_mode, self.cloud_name, self.platform)
+        if strict_mode == "true" and self.cloud_name == CloudNames.UNKNOWN:
             return False
-        elif self.cloud_platform == Platforms.NO_EC2_METADATA:
+        elif self.cloud_name == CloudNames.NO_EC2_METADATA:
             return False
 
         if self.perform_dhcp_setup:  # Setup networking in init-local stage.
@@ -103,13 +90,22 @@ class DataSourceEc2(sources.DataSource):
                 return False
             try:
                 with EphemeralDHCPv4(self.fallback_interface):
-                    return util.log_time(
+                    self._crawled_metadata = util.log_time(
                         logfunc=LOG.debug, msg='Crawl of metadata service',
-                        func=self._crawl_metadata)
+                        func=self.crawl_metadata)
             except NoDHCPLeaseError:
                 return False
         else:
-            return self._crawl_metadata()
+            self._crawled_metadata = util.log_time(
+                logfunc=LOG.debug, msg='Crawl of metadata service',
+                func=self.crawl_metadata)
+        if not self._crawled_metadata:
+            return False
+        self.metadata = self._crawled_metadata.get('meta-data', None)
+        self.userdata_raw = self._crawled_metadata.get('user-data', None)
+        self.identity = self._crawled_metadata.get(
+            'dynamic', {}).get('instance-identity', {}).get('document', {})
+        return True
 
     @property
     def launch_index(self):
@@ -117,6 +113,15 @@ class DataSourceEc2(sources.DataSource):
             return None
         return self.metadata.get('ami-launch-index')
 
+    @property
+    def platform(self):
+        # Handle upgrade path of pickled ds
+        if not hasattr(self, '_platform_type'):
+            self._platform_type = DataSourceEc2.dsname.lower()
+        if not self._platform_type:
+            self._platform_type = DataSourceEc2.dsname.lower()
+        return self._platform_type
+
     def get_metadata_api_version(self):
         """Get the best supported api version from the metadata service.
 
@@ -144,7 +149,7 @@ class DataSourceEc2(sources.DataSource):
         return self.min_metadata_version
 
     def get_instance_id(self):
-        if self.cloud_platform == Platforms.AWS:
+        if self.cloud_name == CloudNames.AWS:
             # Prefer the ID from the instance identity document, but fall back
             if not getattr(self, 'identity', None):
                 # If re-using cached datasource, it's get_data run didn't
@@ -254,7 +259,7 @@ class DataSourceEc2(sources.DataSource):
     @property
     def availability_zone(self):
         try:
-            if self.cloud_platform == Platforms.AWS:
+            if self.cloud_name == CloudNames.AWS:
                 return self.identity.get(
                     'availabilityZone',
                     self.metadata['placement']['availability-zone'])
@@ -265,7 +270,7 @@ class DataSourceEc2(sources.DataSource):
 
     @property
     def region(self):
-        if self.cloud_platform == Platforms.AWS:
+        if self.cloud_name == CloudNames.AWS:
             region = self.identity.get('region')
             # Fallback to trimming the availability zone if region is missing
             if self.availability_zone and not region:
@@ -277,16 +282,10 @@ class DataSourceEc2(sources.DataSource):
                 return az[:-1]
         return None
 
-    @property
-    def cloud_platform(self):  # TODO rename cloud_name
-        if self._cloud_platform is None:
-            self._cloud_platform = identify_platform()
-        return self._cloud_platform
-
     def activate(self, cfg, is_new_instance):
         if not is_new_instance:
             return
-        if self.cloud_platform == Platforms.UNKNOWN:
+        if self.cloud_name == CloudNames.UNKNOWN:
             warn_if_necessary(
                 util.get_cfg_by_path(cfg, STRICT_ID_PATH, STRICT_ID_DEFAULT),
                 cfg)
@@ -306,13 +305,13 @@ class DataSourceEc2(sources.DataSource):
         result = None
         no_network_metadata_on_aws = bool(
             'network' not in self.metadata and
-            self.cloud_platform == Platforms.AWS)
+            self.cloud_name == CloudNames.AWS)
         if no_network_metadata_on_aws:
             LOG.debug("Metadata 'network' not present:"
                       " Refreshing stale metadata from prior to upgrade.")
             util.log_time(
                 logfunc=LOG.debug, msg='Re-crawl of metadata service',
-                func=self._crawl_metadata)
+                func=self.get_data)
 
         # Limit network configuration to only the primary/fallback nic
         iface = self.fallback_interface
@@ -340,28 +339,32 @@ class DataSourceEc2(sources.DataSource):
                 return super(DataSourceEc2, self).fallback_interface
         return self._fallback_interface
 
-    def _crawl_metadata(self):
+    def crawl_metadata(self):
         """Crawl metadata service when available.
 
-        @returns: True on success, False otherwise.
+        @returns: Dictionary of crawled metadata content containing the keys:
+          meta-data, user-data and dynamic.
         """
         if not self.wait_for_metadata_service():
-            return False
+            return {}
         api_version = self.get_metadata_api_version()
+        crawled_metadata = {}
         try:
-            self.userdata_raw = ec2.get_instance_userdata(
+            crawled_metadata['user-data'] = ec2.get_instance_userdata(
                 api_version, self.metadata_address)
-            self.metadata = ec2.get_instance_metadata(
+            crawled_metadata['meta-data'] = ec2.get_instance_metadata(
                 api_version, self.metadata_address)
-            if self.cloud_platform == Platforms.AWS:
-                self.identity = ec2.get_instance_identity(
-                    api_version, self.metadata_address).get('document', {})
+            if self.cloud_name == CloudNames.AWS:
+                identity = ec2.get_instance_identity(
+                    api_version, self.metadata_address)
+                crawled_metadata['dynamic'] = {'instance-identity': identity}
         except Exception:
             util.logexc(
                 LOG, "Failed reading from metadata address %s",
                 self.metadata_address)
-            return False
-        return True
+            return {}
+        crawled_metadata['_metadata_api_version'] = api_version
+        return crawled_metadata
 
 
 class DataSourceEc2Local(DataSourceEc2):
@@ -375,10 +378,10 @@ class DataSourceEc2Local(DataSourceEc2):
     perform_dhcp_setup = True  # Use dhcp before querying metadata
 
     def get_data(self):
-        supported_platforms = (Platforms.AWS,)
-        if self.cloud_platform not in supported_platforms:
+        supported_platforms = (CloudNames.AWS,)
+        if self.cloud_name not in supported_platforms:
             LOG.debug("Local Ec2 mode only supported on %s, not %s",
-                      supported_platforms, self.cloud_platform)
+                      supported_platforms, self.cloud_name)
             return False
         return super(DataSourceEc2Local, self).get_data()
 
@@ -439,20 +442,20 @@ def identify_aws(data):
     if (data['uuid'].startswith('ec2') and
             (data['uuid_source'] == 'hypervisor' or
              data['uuid'] == data['serial'])):
-            return Platforms.AWS
+            return CloudNames.AWS
 
     return None
 
 
 def identify_brightbox(data):
     if data['serial'].endswith('brightbox.com'):
-        return Platforms.BRIGHTBOX
+        return CloudNames.BRIGHTBOX
 
 
 def identify_platform():
-    # identify the platform and return an entry in Platforms.
+    # identify the platform and return an entry in CloudNames.
     data = _collect_platform_data()
-    checks = (identify_aws, identify_brightbox, lambda x: Platforms.UNKNOWN)
+    checks = (identify_aws, identify_brightbox, lambda x: CloudNames.UNKNOWN)
     for checker in checks:
         try:
             result = checker(data)
diff --git a/cloudinit/sources/DataSourceIBMCloud.py b/cloudinit/sources/DataSourceIBMCloud.py
index a535814..21e6ae6 100644
--- a/cloudinit/sources/DataSourceIBMCloud.py
+++ b/cloudinit/sources/DataSourceIBMCloud.py
@@ -157,6 +157,10 @@ class DataSourceIBMCloud(sources.DataSource):
 
         return True
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        return '%s (%s)' % (self.platform, self.source)
+
     def check_instance_id(self, sys_cfg):
         """quickly (local check only) if self.instance_id is still valid
 
diff --git a/cloudinit/sources/DataSourceMAAS.py b/cloudinit/sources/DataSourceMAAS.py
index bcb3854..61aa6d7 100644
--- a/cloudinit/sources/DataSourceMAAS.py
+++ b/cloudinit/sources/DataSourceMAAS.py
@@ -109,6 +109,10 @@ class DataSourceMAAS(sources.DataSource):
                 LOG.warning("Invalid content in vendor-data: %s", e)
                 self.vendordata_raw = None
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        return 'seed-dir (%s)' % self.base_url
+
     def wait_for_metadata_service(self, url):
         mcfg = self.ds_cfg
         max_wait = 120
diff --git a/cloudinit/sources/DataSourceNoCloud.py b/cloudinit/sources/DataSourceNoCloud.py
index 2daea59..9010f06 100644
--- a/cloudinit/sources/DataSourceNoCloud.py
+++ b/cloudinit/sources/DataSourceNoCloud.py
@@ -186,6 +186,27 @@ class DataSourceNoCloud(sources.DataSource):
         self._network_eni = mydata['meta-data'].get('network-interfaces')
         return True
 
+    @property
+    def platform_type(self):
+        # Handle upgrade path of pickled ds
+        if not hasattr(self, '_platform_type'):
+            self._platform_type = None
+        if not self._platform_type:
+            self._platform_type = 'lxd' if util.is_lxd() else 'nocloud'
+        return self._platform_type
+
+    def _get_cloud_name(self):
+        """Return unknown when 'cloud-name' key is absent from metadata."""
+        return sources.METADATA_UNKNOWN
+
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        if self.seed.startswith('/dev'):
+            subplatform_type = 'config-disk'
+        else:
+            subplatform_type = 'seed-dir'
+        return '%s (%s)' % (subplatform_type, self.seed)
+
     def check_instance_id(self, sys_cfg):
         # quickly (local check only) if self.instance_id is still valid
         # we check kernel command line or files.
diff --git a/cloudinit/sources/DataSourceNone.py b/cloudinit/sources/DataSourceNone.py
index e63a7e3..e625080 100644
--- a/cloudinit/sources/DataSourceNone.py
+++ b/cloudinit/sources/DataSourceNone.py
@@ -28,6 +28,10 @@ class DataSourceNone(sources.DataSource):
             self.metadata = self.ds_cfg['metadata']
         return True
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        return 'config'
+
     def get_instance_id(self):
         return 'iid-datasource-none'
 
diff --git a/cloudinit/sources/DataSourceOVF.py b/cloudinit/sources/DataSourceOVF.py
index 178ccb0..045291e 100644
--- a/cloudinit/sources/DataSourceOVF.py
+++ b/cloudinit/sources/DataSourceOVF.py
@@ -275,6 +275,12 @@ class DataSourceOVF(sources.DataSource):
         self.cfg = cfg
         return True
 
+    def _get_subplatform(self):
+        system_type = util.read_dmi_data("system-product-name").lower()
+        if system_type == 'vmware':
+            return 'vmware (%s)' % self.seed
+        return 'ovf (%s)' % self.seed
+
     def get_public_ssh_keys(self):
         if 'public-keys' not in self.metadata:
             return []
diff --git a/cloudinit/sources/DataSourceOpenNebula.py b/cloudinit/sources/DataSourceOpenNebula.py
index 77ccd12..e62e972 100644
--- a/cloudinit/sources/DataSourceOpenNebula.py
+++ b/cloudinit/sources/DataSourceOpenNebula.py
@@ -95,6 +95,14 @@ class DataSourceOpenNebula(sources.DataSource):
         self.userdata_raw = results.get('userdata')
         return True
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        if self.seed_dir in self.seed:
+            subplatform_type = 'seed-dir'
+        else:
+            subplatform_type = 'config-disk'
+        return '%s (%s)' % (subplatform_type, self.seed)
+
     @property
     def network_config(self):
         if self.network is not None:
diff --git a/cloudinit/sources/DataSourceOracle.py b/cloudinit/sources/DataSourceOracle.py
index fab39af..70b9c58 100644
--- a/cloudinit/sources/DataSourceOracle.py
+++ b/cloudinit/sources/DataSourceOracle.py
@@ -91,6 +91,10 @@ class DataSourceOracle(sources.DataSource):
     def crawl_metadata(self):
         return read_metadata()
 
+    def _get_subplatform(self):
+        """Return the subplatform metadata source details."""
+        return 'metadata (%s)' % METADATA_ENDPOINT
+
     def check_instance_id(self, sys_cfg):
         """quickly check (local only) if self.instance_id is still valid
 
diff --git a/cloudinit/sources/DataSourceSmartOS.py b/cloudinit/sources/DataSourceSmartOS.py
index 593ac91..32b57cd 100644
--- a/cloudinit/sources/DataSourceSmartOS.py
+++ b/cloudinit/sources/DataSourceSmartOS.py
@@ -303,6 +303,9 @@ class DataSourceSmartOS(sources.DataSource):
         self._set_provisioned()
         return True
 
+    def _get_subplatform(self):
+        return 'serial (%s)' % SERIAL_DEVICE
+
     def device_name_to_device(self, name):
         return self.ds_cfg['disk_aliases'].get(name)
 
diff --git a/cloudinit/sources/__init__.py b/cloudinit/sources/__init__.py
index 5ac9882..2cf76ad 100644
--- a/cloudinit/sources/__init__.py
+++ b/cloudinit/sources/__init__.py
@@ -54,9 +54,20 @@ REDACT_SENSITIVE_VALUE = 'redacted for non-root user'
 METADATA_CLOUD_NAME_KEY = 'cloud-name'
 
 UNSET = "_unset"
+METADATA_UNKNOWN = 'unknown'
 
 LOG = logging.getLogger(__name__)
 
+# CLOUD_ID_REGION_PREFIX_MAP format is:
+#  <region-match-prefix>: (<new-cloud-id>: <test_allowed_cloud_callable>)
+CLOUD_ID_REGION_PREFIX_MAP = {
+    'cn-': ('aws-china', lambda c, _: c == 'aws'),    # only change aws regions
+    'us-gov-': ('aws-gov', lambda c, _: c == 'aws'),  # only change aws regions
+    'china': ('azure-china', lambda c, _: c == 'azure'),  # only change azure
+    ('dfw', 'ord', 'iad', 'lon', 'syd', 'hkg'): (
+         'rackspace' , lambda c, p: (c == 'unknown' and p == 'openstack'))
+}
+
 
 class DataSourceNotFoundException(Exception):
     pass
@@ -133,6 +144,14 @@ class DataSource(object):
     # Cached cloud_name as determined by _get_cloud_name
     _cloud_name = None
 
+    # Cached cloud platform api type: e.g. ec2, openstack, kvm, lxd, azure etc.
+    _platform_type = None
+
+    # More details about the cloud platform:
+    #  - metadata (http://169.254.169.254/)
+    #  - seed-dir (<dirname>)
+    _subplatform = None
+
     # Track the discovered fallback nic for use in configuration generation.
     _fallback_interface = None
 
@@ -192,21 +211,24 @@ class DataSource(object):
         local_hostname = self.get_hostname()
         instance_id = self.get_instance_id()
         availability_zone = self.availability_zone
-        cloud_name = self.cloud_name
-        # When adding new standard keys prefer underscore-delimited instead
-        # of hyphen-delimted to support simple variable references in jinja
-        # templates.
+        # In the event of upgrade from existing cloudinit, pickled datasource
+        # will not contain these new class attributes. So we need to recrawl
+        # metadata to discover that content.
         return {
             'v1': {
+                '_beta_keys': ['subplatform'],
                 'availability-zone': availability_zone,
                 'availability_zone': availability_zone,
-                'cloud-name': cloud_name,
-                'cloud_name': cloud_name,
+                'cloud-name': self.cloud_name,
+                'cloud_name': self.cloud_name,
+                'platform': self.platform_type,
+                'public_ssh_keys': self.get_public_ssh_keys(),
                 'instance-id': instance_id,
                 'instance_id': instance_id,
                 'local-hostname': local_hostname,
                 'local_hostname': local_hostname,
-                'region': self.region}}
+                'region': self.region,
+                'subplatform': self.subplatform}}
 
     def clear_cached_attrs(self, attr_defaults=()):
         """Reset any cached metadata attributes to datasource defaults.
@@ -247,19 +269,27 @@ class DataSource(object):
 
         @return True on successful write, False otherwise.
         """
-        instance_data = {
-            'ds': {'_doc': EXPERIMENTAL_TEXT,
-                   'meta_data': self.metadata}}
-        if hasattr(self, 'network_json'):
-            network_json = getattr(self, 'network_json')
-            if network_json != UNSET:
-                instance_data['ds']['network_json'] = network_json
-        if hasattr(self, 'ec2_metadata'):
-            ec2_metadata = getattr(self, 'ec2_metadata')
-            if ec2_metadata != UNSET:
-                instance_data['ds']['ec2_metadata'] = ec2_metadata
+        if hasattr(self, '_crawled_metadata'):
+            # Any datasource with _crawled_metadata will best represent
+            # most recent, 'raw' metadata
+            crawled_metadata = copy.deepcopy(
+                getattr(self, '_crawled_metadata'))
+            crawled_metadata.pop('user-data', None)
+            crawled_metadata.pop('vendor-data', None)
+            instance_data = {'ds': crawled_metadata}
+        else:
+            instance_data = {'ds': {'meta_data': self.metadata}}
+            if hasattr(self, 'network_json'):
+                network_json = getattr(self, 'network_json')
+                if network_json != UNSET:
+                    instance_data['ds']['network_json'] = network_json
+            if hasattr(self, 'ec2_metadata'):
+                ec2_metadata = getattr(self, 'ec2_metadata')
+                if ec2_metadata != UNSET:
+                    instance_data['ds']['ec2_metadata'] = ec2_metadata
         instance_data.update(
             self._get_standardized_metadata())
+        instance_data['ds']['_doc'] = EXPERIMENTAL_TEXT
         try:
             # Process content base64encoding unserializable values
             content = util.json_dumps(instance_data)
@@ -347,6 +377,40 @@ class DataSource(object):
         return self._fallback_interface
 
     @property
+    def platform_type(self):
+        if not hasattr(self, '_platform_type'):
+            # Handle upgrade path where pickled datasource has no _platform.
+            self._platform_type = self.dsname.lower()
+        if not self._platform_type:
+            self._platform_type = self.dsname.lower()
+        return self._platform_type
+
+    @property
+    def subplatform(self):
+        """Return a string representing subplatform details for the datasource.
+
+        This should be guidance for where the metadata is sourced.
+        Examples of this on different clouds:
+            ec2:       metadata (http://169.254.169.254)
+            openstack: configdrive (/dev/path)
+            openstack: metadata (http://169.254.169.254)
+            nocloud:   seed-dir (/seed/dir/path)
+            lxd:   nocloud (/seed/dir/path)
+        """
+        if not hasattr(self, '_subplatform'):
+            # Handle upgrade path where pickled datasource has no _platform.
+            self._subplatform = self._get_subplatform()
+        if not self._subplatform:
+            self._subplatform = self._get_subplatform()
+        return self._subplatform
+
+    def _get_subplatform(self):
+        """Subclasses should implement to return a "slug (detail)" string."""
+        if hasattr(self, 'metadata_address'):
+            return 'metadata (%s)' % getattr(self, 'metadata_address')
+        return METADATA_UNKNOWN
+
+    @property
     def cloud_name(self):
         """Return lowercase cloud name as determined by the datasource.
 
@@ -359,9 +423,11 @@ class DataSource(object):
             cloud_name = self.metadata.get(METADATA_CLOUD_NAME_KEY)
             if isinstance(cloud_name, six.string_types):
                 self._cloud_name = cloud_name.lower()
-            LOG.debug(
-                'Ignoring metadata provided key %s: non-string type %s',
-                METADATA_CLOUD_NAME_KEY, type(cloud_name))
+            else:
+                self._cloud_name = self._get_cloud_name().lower()
+                LOG.debug(
+                    'Ignoring metadata provided key %s: non-string type %s',
+                    METADATA_CLOUD_NAME_KEY, type(cloud_name))
         else:
             self._cloud_name = self._get_cloud_name().lower()
         return self._cloud_name
@@ -714,6 +780,26 @@ def instance_id_matches_system_uuid(instance_id, field='system-uuid'):
     return instance_id.lower() == dmi_value.lower()
 
 
+def canonical_cloud_id(cloud_name, region, platform):
+    """Lookup the canonical cloud-id for a given cloud_name and region."""
+    if not region or region == METADATA_UNKNOWN:
+        if cloud_name != METADATA_UNKNOWN:
+            return cloud_name
+        return platform
+    for prefix, cloud_id_test in CLOUD_ID_REGION_PREFIX_MAP.items():
+        (cloud_id, valid_cloud) = cloud_id_test
+        if isinstance(prefix, tuple):
+            for item in prefix:
+                if item in region and valid_cloud(cloud_name, platform):
+                    return cloud_id
+        else:
+            if prefix in region and valid_cloud(cloud_name, platform):
+                return cloud_id
+    if cloud_name != METADATA_UNKNOWN:
+        return cloud_name
+    return platform
+
+
 def convert_vendordata(data, recurse=True):
     """data: a loaded object (strings, arrays, dicts).
     return something suitable for cloudinit vendordata_raw.
diff --git a/cloudinit/sources/tests/test_init.py b/cloudinit/sources/tests/test_init.py
index 8082019..391b343 100644
--- a/cloudinit/sources/tests/test_init.py
+++ b/cloudinit/sources/tests/test_init.py
@@ -295,6 +295,7 @@ class TestDataSource(CiTestCase):
             'base64_encoded_keys': [],
             'sensitive_keys': [],
             'v1': {
+                '_beta_keys': ['subplatform'],
                 'availability-zone': 'myaz',
                 'availability_zone': 'myaz',
                 'cloud-name': 'subclasscloudname',
@@ -303,7 +304,10 @@ class TestDataSource(CiTestCase):
                 'instance_id': 'iid-datasource',
                 'local-hostname': 'test-subclass-hostname',
                 'local_hostname': 'test-subclass-hostname',
-                'region': 'myregion'},
+                'platform': 'mytestsubclass',
+                'public_ssh_keys': [],
+                'region': 'myregion',
+                'subplatform': 'unknown'},
             'ds': {
                 '_doc': EXPERIMENTAL_TEXT,
                 'meta_data': {'availability_zone': 'myaz',
@@ -339,6 +343,7 @@ class TestDataSource(CiTestCase):
             'base64_encoded_keys': [],
             'sensitive_keys': ['ds/meta_data/some/security-credentials'],
             'v1': {
+                '_beta_keys': ['subplatform'],
                 'availability-zone': 'myaz',
                 'availability_zone': 'myaz',
                 'cloud-name': 'subclasscloudname',
@@ -347,7 +352,10 @@ class TestDataSource(CiTestCase):
                 'instance_id': 'iid-datasource',
                 'local-hostname': 'test-subclass-hostname',
                 'local_hostname': 'test-subclass-hostname',
-                'region': 'myregion'},
+                'platform': 'mytestsubclass',
+                'public_ssh_keys': [],
+                'region': 'myregion',
+                'subplatform': 'unknown'},
             'ds': {
                 '_doc': EXPERIMENTAL_TEXT,
                 'meta_data': {
diff --git a/cloudinit/sources/tests/test_oracle.py b/cloudinit/sources/tests/test_oracle.py
index 7599126..97d6294 100644
--- a/cloudinit/sources/tests/test_oracle.py
+++ b/cloudinit/sources/tests/test_oracle.py
@@ -71,6 +71,14 @@ class TestDataSourceOracle(test_helpers.CiTestCase):
         self.assertFalse(ds._get_data())
         mocks._is_platform_viable.assert_called_once_with()
 
+    def test_platform_info(self):
+        """Return platform-related information for Oracle Datasource."""
+        ds, _mocks = self._get_ds()
+        self.assertEqual('oracle', ds.cloud_name)
+        self.assertEqual('oracle', ds.platform_type)
+        self.assertEqual(
+            'metadata (http://169.254.169.254/openstack/)', ds.subplatform)
+
     @mock.patch(DS_PATH + "._is_iscsi_root", return_value=True)
     def test_without_userdata(self, m_is_iscsi_root):
         """If no user-data is provided, it should not be in return dict."""
diff --git a/cloudinit/tests/test_util.py b/cloudinit/tests/test_util.py
index edb0c18..749a384 100644
--- a/cloudinit/tests/test_util.py
+++ b/cloudinit/tests/test_util.py
@@ -478,4 +478,20 @@ class TestGetLinuxDistro(CiTestCase):
         dist = util.get_linux_distro()
         self.assertEqual(('foo', '1.1', 'aarch64'), dist)
 
+
[e-mail]('os.path.exists')
+class TestIsLXD(CiTestCase):
+
+    def test_is_lxd_true_on_sock_device(self, m_exists):
+        """When lxd's /dev/lxd/sock exists, is_lxd returns true."""
+        m_exists.return_value = True
+        self.assertTrue(util.is_lxd())
+        m_exists.assert_called_once_with('/dev/lxd/sock')
+
+    def test_is_lxd_false_when_sock_device_absent(self, m_exists):
+        """When lxd's /dev/lxd/sock is absent, is_lxd returns false."""
+        m_exists.return_value = False
+        self.assertFalse(util.is_lxd())
+        m_exists.assert_called_once_with('/dev/lxd/sock')
+
 # vi: ts=4 expandtab
diff --git a/cloudinit/util.py b/cloudinit/util.py
index 5068096..c67d6be 100644
--- a/cloudinit/util.py
+++ b/cloudinit/util.py
@@ -2171,6 +2171,11 @@ def is_container():
     return False
 
 
+def is_lxd():
+    """Check to see if we are running in a lxd container."""
+    return os.path.exists('/dev/lxd/sock')
+
+
 def get_proc_env(pid, encoding='utf-8', errors='replace'):
     """
     Return the environment in a dict that a given process id was started with.
diff --git a/doc/rtd/topics/instancedata.rst b/doc/rtd/topics/instancedata.rst
index 634e180..5d2dc94 100644
--- a/doc/rtd/topics/instancedata.rst
+++ b/doc/rtd/topics/instancedata.rst
@@ -90,24 +90,46 @@ There are three basic top-level keys:
 
 The standardized keys present:
 
-+----------------------+-----------------------------------------------+---------------------------+
-|  Key path            | Description                                   | Examples                  |
-+======================+===============================================+===========================+
-| v1.cloud_name        | The name of the cloud provided by metadata    | aws, openstack, azure,    |
-|                      | key 'cloud-name' or the cloud-init datasource | configdrive, nocloud,     |
-|                      | name which was discovered.                    | ovf, etc.                 |
-+----------------------+-----------------------------------------------+---------------------------+
-| v1.instance_id       | Unique instance_id allocated by the cloud     | i-<somehash>              |
-+----------------------+-----------------------------------------------+---------------------------+
-| v1.local_hostname    | The internal or local hostname of the system  | ip-10-41-41-70,           |
-|                      |                                               | <user-provided-hostname>  |
-+----------------------+-----------------------------------------------+---------------------------+
-| v1.region            | The physical region/datacenter in which the   | us-east-2                 |
-|                      | instance is deployed                          |                           |
-+----------------------+-----------------------------------------------+---------------------------+
-| v1.availability_zone | The physical availability zone in which the   | us-east-2b, nova, null    |
-|                      | instance is deployed                          |                           |
-+----------------------+-----------------------------------------------+---------------------------+
++----------------------+-----------------------------------------------+-----------------------------------+
+|  Key path            | Description                                   | Examples                          |
++======================+===============================================+===================================+
+| v1._beta_keys        | List of standardized keys still in 'beta'.    | [subplatform]                     |
+|                      | The format, intent or presence of these keys  |                                   |
+|                      | can change. Do not consider them              |                                   |
+|                      | production-ready.                             |                                   |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.cloud_name        | Where possible this will indicate the 'name'  | aws, openstack, azure,            |
+|                      | of the cloud this system is running on.  This | configdrive, nocloud,             |
+|                      | is specifically different than the 'platform' | ovf, etc.                         |
+|                      | below.  As an example, the name of Amazon Web |                                   |
+|                      | Services is 'aws' while the platform is 'ec2'.|                                   |
+|                      |                                               |                                   |
+|                      | If no specific name is determinable or        |                                   |
+|                      | provided in meta-data, then this field may    |                                   |
+|                      | contain the same content as 'platform'.       |                                   |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.instance_id       | Unique instance_id allocated by the cloud     | i-<somehash>                      |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.local_hostname    | The internal or local hostname of the system  | ip-10-41-41-70,                   |
+|                      |                                               | <user-provided-hostname>          |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.platform          | An attempt to identify the cloud platform     | ec2, openstack, lxd, gce          |
+|                      | instance that the system is running on.       | nocloud, ovf                      |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.subplatform       | Additional platform details describing the    | metadata (http://168.254.169.254),|
+|                      | specific source or type of metadata used.     | seed-dir (/path/to/seed-dir/),    |
+|                      | The format of subplatform will be:            | config-disk (/dev/cd0),           |
+|                      | <subplatform_type> (<url_file_or_dev_path>)   | configdrive (/dev/sr0)            |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.public_ssh_keys   | A list of  ssh keys provided to the instance  | ['ssh-rsa AA...', ...]            |
+|                      | by the datasource metadata.                   |                                   |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.region            | The physical region/datacenter in which the   | us-east-2                         |
+|                      | instance is deployed                          |                                   |
++----------------------+-----------------------------------------------+-----------------------------------+
+| v1.availability_zone | The physical availability zone in which the   | us-east-2b, nova, null            |
+|                      | instance is deployed                          |                                   |
++----------------------+-----------------------------------------------+-----------------------------------+
 
 
 Below is an example of ``/run/cloud-init/instance_data.json`` on an EC2
@@ -117,10 +139,75 @@ instance:
 
   {
    "base64_encoded_keys": [],
-   "sensitive_keys": [],
    "ds": {
-    "meta_data": {
-     "ami-id": "ami-014e1416b628b0cbf",
+    "_doc": "EXPERIMENTAL: The structure and format of content scoped under the 'ds' key may change in subsequent releases of cloud-init.",
+    "_metadata_api_version": "2016-09-02",
+    "dynamic": {
+     "instance-identity": {
+      "document": {
+       "accountId": "437526006925",
+       "architecture": "x86_64",
+       "availabilityZone": "us-east-2b",
+       "billingProducts": null,
+       "devpayProductCodes": null,
+       "imageId": "ami-079638aae7046bdd2",
+       "instanceId": "i-075f088c72ad3271c",
+       "instanceType": "t2.micro",
+       "kernelId": null,
+       "marketplaceProductCodes": null,
+       "pendingTime": "2018-10-05T20:10:43Z",
+       "privateIp": "10.41.41.95",
+       "ramdiskId": null,
+       "region": "us-east-2",
+       "version": "2017-09-30"
+      },
+      "pkcs7": [
+       "MIAGCSqGSIb3DQEHAqCAMIACAQExCzAJBgUrDgMCGgUAMIAGCSqGSIb3DQEHAaCAJIAEggHbewog",
+       "ICJkZXZwYXlQcm9kdWN0Q29kZXMiIDogbnVsbCwKICAibWFya2V0cGxhY2VQcm9kdWN0Q29kZXMi",
+       "IDogbnVsbCwKICAicHJpdmF0ZUlwIiA6ICIxMC40MS40MS45NSIsCiAgInZlcnNpb24iIDogIjIw",
+       "MTctMDktMzAiLAogICJpbnN0YW5jZUlkIiA6ICJpLTA3NWYwODhjNzJhZDMyNzFjIiwKICAiYmls",
+       "bGluZ1Byb2R1Y3RzIiA6IG51bGwsCiAgImluc3RhbmNlVHlwZSIgOiAidDIubWljcm8iLAogICJh",
+       "Y2NvdW50SWQiIDogIjQzNzUyNjAwNjkyNSIsCiAgImF2YWlsYWJpbGl0eVpvbmUiIDogInVzLWVh",
+       "c3QtMmIiLAogICJrZXJuZWxJZCIgOiBudWxsLAogICJyYW1kaXNrSWQiIDogbnVsbCwKICAiYXJj",
+       "aGl0ZWN0dXJlIiA6ICJ4ODZfNjQiLAogICJpbWFnZUlkIiA6ICJhbWktMDc5NjM4YWFlNzA0NmJk",
+       "ZDIiLAogICJwZW5kaW5nVGltZSIgOiAiMjAxOC0xMC0wNVQyMDoxMDo0M1oiLAogICJyZWdpb24i",
+       "IDogInVzLWVhc3QtMiIKfQAAAAAAADGCARcwggETAgEBMGkwXDELMAkGA1UEBhMCVVMxGTAXBgNV",
+       "BAgTEFdhc2hpbmd0b24gU3RhdGUxEDAOBgNVBAcTB1NlYXR0bGUxIDAeBgNVBAoTF0FtYXpvbiBX",
+       "ZWIgU2VydmljZXMgTExDAgkAlrpI2eVeGmcwCQYFKw4DAhoFAKBdMBgGCSqGSIb3DQEJAzELBgkq",
+       "hkiG9w0BBwEwHAYJKoZIhvcNAQkFMQ8XDTE4MTAwNTIwMTA0OFowIwYJKoZIhvcNAQkEMRYEFK0k",
+       "Tz6n1A8/zU1AzFj0riNQORw2MAkGByqGSM44BAMELjAsAhRNrr174y98grPBVXUforN/6wZp8AIU",
+       "JLZBkrB2GJA8A4WJ1okq++jSrBIAAAAAAAA="
+      ],
+      "rsa2048": [
+       "MIAGCSqGSIb3DQEHAqCAMIACAQExDzANBglghkgBZQMEAgEFADCABgkqhkiG9w0BBwGggCSABIIB",
+       "23sKICAiZGV2cGF5UHJvZHVjdENvZGVzIiA6IG51bGwsCiAgIm1hcmtldHBsYWNlUHJvZHVjdENv",
+       "ZGVzIiA6IG51bGwsCiAgInByaXZhdGVJcCIgOiAiMTAuNDEuNDEuOTUiLAogICJ2ZXJzaW9uIiA6",
+       "ICIyMDE3LTA5LTMwIiwKICAiaW5zdGFuY2VJZCIgOiAiaS0wNzVmMDg4YzcyYWQzMjcxYyIsCiAg",
+       "ImJpbGxpbmdQcm9kdWN0cyIgOiBudWxsLAogICJpbnN0YW5jZVR5cGUiIDogInQyLm1pY3JvIiwK",
+       "ICAiYWNjb3VudElkIiA6ICI0Mzc1MjYwMDY5MjUiLAogICJhdmFpbGFiaWxpdHlab25lIiA6ICJ1",
+       "cy1lYXN0LTJiIiwKICAia2VybmVsSWQiIDogbnVsbCwKICAicmFtZGlza0lkIiA6IG51bGwsCiAg",
+       "ImFyY2hpdGVjdHVyZSIgOiAieDg2XzY0IiwKICAiaW1hZ2VJZCIgOiAiYW1pLTA3OTYzOGFhZTcw",
+       "NDZiZGQyIiwKICAicGVuZGluZ1RpbWUiIDogIjIwMTgtMTAtMDVUMjA6MTA6NDNaIiwKICAicmVn",
+       "aW9uIiA6ICJ1cy1lYXN0LTIiCn0AAAAAAAAxggH/MIIB+wIBATBpMFwxCzAJBgNVBAYTAlVTMRkw",
+       "FwYDVQQIExBXYXNoaW5ndG9uIFN0YXRlMRAwDgYDVQQHEwdTZWF0dGxlMSAwHgYDVQQKExdBbWF6",
+       "b24gV2ViIFNlcnZpY2VzIExMQwIJAM07oeX4xevdMA0GCWCGSAFlAwQCAQUAoGkwGAYJKoZIhvcN",
+       "AQkDMQsGCSqGSIb3DQEHATAcBgkqhkiG9w0BCQUxDxcNMTgxMDA1MjAxMDQ4WjAvBgkqhkiG9w0B",
+       "CQQxIgQgkYz0pZk3zJKBi4KP4egeOKJl/UYwu5UdE7id74pmPwMwDQYJKoZIhvcNAQEBBQAEggEA",
+       "dC3uIGGNul1OC1mJKSH3XoBWsYH20J/xhIdftYBoXHGf2BSFsrs9ZscXd2rKAKea4pSPOZEYMXgz",
+       "lPuT7W0WU89N3ZKviy/ReMSRjmI/jJmsY1lea6mlgcsJXreBXFMYucZvyeWGHdnCjamoKWXkmZlM",
+       "mSB1gshWy8Y7DzoKviYPQZi5aI54XK2Upt4kGme1tH1NI2Cq+hM4K+adxTbNhS3uzvWaWzMklUuU",
+       "QHX2GMmjAVRVc8vnA8IAsBCJJp+gFgYzi09IK+cwNgCFFPADoG6jbMHHf4sLB3MUGpiA+G9JlCnM",
+       "fmkjI2pNRB8spc0k4UG4egqLrqCz67WuK38tjwAAAAAAAA=="
+      ],
+      "signature": [
+       "Tsw6h+V3WnxrNVSXBYIOs1V4j95YR1mLPPH45XnhX0/Ei3waJqf7/7EEKGYP1Cr4PTYEULtZ7Mvf",
+       "+xJpM50Ivs2bdF7o0c4vnplRWe3f06NI9pv50dr110j/wNzP4MZ1pLhJCqubQOaaBTF3LFutgRrt",
+       "r4B0mN3p7EcqD8G+ll0="
+      ]
+     }
+    },
+    "meta-data": {
+     "ami-id": "ami-079638aae7046bdd2",
      "ami-launch-index": "0",
      "ami-manifest-path": "(unknown)",
      "block-device-mapping": {
@@ -129,31 +216,31 @@ instance:
       "ephemeral1": "sdc",
       "root": "/dev/sda1"
      },
-     "hostname": "ip-10-41-41-70.us-east-2.compute.internal",
+     "hostname": "ip-10-41-41-95.us-east-2.compute.internal",
      "instance-action": "none",
-     "instance-id": "i-04fa31cfc55aa7976",
+     "instance-id": "i-075f088c72ad3271c",
      "instance-type": "t2.micro",
-     "local-hostname": "ip-10-41-41-70.us-east-2.compute.internal",
-     "local-ipv4": "10.41.41.70",
-     "mac": "06:b6:92:dd:9d:24",
+     "local-hostname": "ip-10-41-41-95.us-east-2.compute.internal",
+     "local-ipv4": "10.41.41.95",
+     "mac": "06:74:8f:39:cd:a6",
      "metrics": {
       "vhostmd": "<?xml version=\"1.0\" encoding=\"UTF-8\"?>"
      },
      "network": {
       "interfaces": {
        "macs": {
-	"06:b6:92:dd:9d:24": {
+	"06:74:8f:39:cd:a6": {
 	 "device-number": "0",
-	 "interface-id": "eni-08c0c9fdb99b6e6f4",
+	 "interface-id": "eni-052058bbd7831eaae",
 	 "ipv4-associations": {
-	  "18.224.22.43": "10.41.41.70"
+	  "18.218.221.122": "10.41.41.95"
 	 },
-	 "local-hostname": "ip-10-41-41-70.us-east-2.compute.internal",
-	 "local-ipv4s": "10.41.41.70",
-	 "mac": "06:b6:92:dd:9d:24",
+	 "local-hostname": "ip-10-41-41-95.us-east-2.compute.internal",
+	 "local-ipv4s": "10.41.41.95",
+	 "mac": "06:74:8f:39:cd:a6",
 	 "owner-id": "437526006925",
-	 "public-hostname": "ec2-18-224-22-43.us-east-2.compute.amazonaws.com",
-	 "public-ipv4s": "18.224.22.43",
+	 "public-hostname": "ec2-18-218-221-122.us-east-2.compute.amazonaws.com",
+	 "public-ipv4s": "18.218.221.122",
 	 "security-group-ids": "sg-828247e9",
 	 "security-groups": "Cloud-init integration test secgroup",
 	 "subnet-id": "subnet-282f3053",
@@ -171,16 +258,14 @@ instance:
       "availability-zone": "us-east-2b"
      },
      "profile": "default-hvm",
-     "public-hostname": "ec2-18-224-22-43.us-east-2.compute.amazonaws.com",
-     "public-ipv4": "18.224.22.43",
+     "public-hostname": "ec2-18-218-221-122.us-east-2.compute.amazonaws.com",
+     "public-ipv4": "18.218.221.122",
      "public-keys": {
       "cloud-init-integration": [
-       "ssh-rsa
-  AAAAB3NzaC1yc2EAAAADAQABAAABAQDSL7uWGj8cgWyIOaspgKdVy0cKJ+UTjfv7jBOjG2H/GN8bJVXy72XAvnhM0dUM+CCs8FOf0YlPX+Frvz2hKInrmRhZVwRSL129PasD12MlI3l44u6IwS1o/W86Q+tkQYEljtqDOo0a+cOsaZkvUNzUyEXUwz/lmYa6G4hMKZH4NBj7nbAAF96wsMCoyNwbWryBnDYUr6wMbjRR1J9Pw7Xh7WRC73wy4Va2YuOgbD3V/5ZrFPLbWZW/7TFXVrql04QVbyei4aiFR5n//GvoqwQDNe58LmbzX/xvxyKJYdny2zXmdAhMxbrpFQsfpkJ9E/H5w0yOdSvnWbUoG5xNGoOB
-  cloud-init-integration"
+       "ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQDSL7uWGj8cgWyIOaspgKdVy0cKJ+UTjfv7jBOjG2H/GN8bJVXy72XAvnhM0dUM+CCs8FOf0YlPX+Frvz2hKInrmRhZVwRSL129PasD12MlI3l44u6IwS1o/W86Q+tkQYEljtqDOo0a+cOsaZkvUNzUyEXUwz/lmYa6G4hMKZH4NBj7nbAAF96wsMCoyNwbWryBnDYUr6wMbjRR1J9Pw7Xh7WRC73wy4Va2YuOgbD3V/5ZrFPLbWZW/7TFXVrql04QVbyei4aiFR5n//GvoqwQDNe58LmbzX/xvxyKJYdny2zXmdAhMxbrpFQsfpkJ9E/H5w0yOdSvnWbUoG5xNGoOB cloud-init-integration"
       ]
      },
-     "reservation-id": "r-06ab75e9346f54333",
+     "reservation-id": "r-0594a20e31f6cfe46",
      "security-groups": "Cloud-init integration test secgroup",
      "services": {
       "domain": "amazonaws.com",
@@ -188,16 +273,22 @@ instance:
      }
     }
    },
+   "sensitive_keys": [],
    "v1": {
+    "_beta_keys": [
+     "subplatform"
+    ],
     "availability-zone": "us-east-2b",
     "availability_zone": "us-east-2b",
-    "cloud-name": "aws",
     "cloud_name": "aws",
-    "instance-id": "i-04fa31cfc55aa7976",
-    "instance_id": "i-04fa31cfc55aa7976",
-    "local-hostname": "ip-10-41-41-70",
-    "local_hostname": "ip-10-41-41-70",
-    "region": "us-east-2"
+    "instance_id": "i-075f088c72ad3271c",
+    "local_hostname": "ip-10-41-41-95",
+    "platform": "ec2",
+    "public_ssh_keys": [
+     "ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAABAQDSL7uWGj8cgWyIOaspgKdVy0cKJ+UTjfv7jBOjG2H/GN8bJVXy72XAvnhM0dUM+CCs8FOf0YlPX+Frvz2hKInrmRhZVwRSL129PasD12MlI3l44u6IwS1o/W86Q+tkQYEljtqDOo0a+cOsaZkvUNzUyEXUwz/lmYa6G4hMKZH4NBj7nbAAF96wsMCoyNwbWryBnDYUr6wMbjRR1J9Pw7Xh7WRC73wy4Va2YuOgbD3V/5ZrFPLbWZW/7TFXVrql04QVbyei4aiFR5n//GvoqwQDNe58LmbzX/xvxyKJYdny2zXmdAhMxbrpFQsfpkJ9E/H5w0yOdSvnWbUoG5xNGoOB cloud-init-integration"
+    ],
+    "region": "us-east-2",
+    "subplatform": "metadata (http://169.254.169.254)"
    }
   }
 
diff --git a/setup.py b/setup.py
index 5ed8eae..ea37efc 100755
--- a/setup.py
+++ b/setup.py
@@ -282,7 +282,8 @@ setuptools.setup(
     cmdclass=cmdclass,
     entry_points={
         'console_scripts': [
-            'cloud-init = cloudinit.cmd.main:main'
+            'cloud-init = cloudinit.cmd.main:main',
+            'cloud-id = cloudinit.cmd.cloud_id:main'
         ],
     }
 )
diff --git a/tests/cloud_tests/testcases/base.py b/tests/cloud_tests/testcases/base.py
index e18d601..16b268e 100644
--- a/tests/cloud_tests/testcases/base.py
+++ b/tests/cloud_tests/testcases/base.py
@@ -195,6 +195,9 @@ class CloudTestCase(unittest2.TestCase):
         self.assertIsNotNone(
             v1_data['availability_zone'], 'expected ec2 availability_zone')
         self.assertEqual('aws', v1_data['cloud_name'])
+        self.assertEqual('ec2', v1_data['platform'])
+        self.assertEqual(
+            'metadata (http://169.254.169.254)', v1_data['subplatform'])
         self.assertIn('i-', v1_data['instance_id'])
         self.assertIn('ip-', v1_data['local_hostname'])
         self.assertIsNotNone(v1_data['region'], 'expected ec2 region')
@@ -220,7 +223,11 @@ class CloudTestCase(unittest2.TestCase):
         instance_data = json.loads(out)
         v1_data = instance_data.get('v1', {})
         self.assertItemsEqual([], sorted(instance_data['base64_encoded_keys']))
-        self.assertEqual('nocloud', v1_data['cloud_name'])
+        self.assertEqual('unknown', v1_data['cloud_name'])
+        self.assertEqual('lxd', v1_data['platform'])
+        self.assertEqual(
+            'seed-dir (/var/lib/cloud/seed/nocloud-net)',
+            v1_data['subplatform'])
         self.assertIsNone(
             v1_data['availability_zone'],
             'found unexpected lxd availability_zone %s' %
@@ -253,7 +260,9 @@ class CloudTestCase(unittest2.TestCase):
         instance_data = json.loads(out)
         v1_data = instance_data.get('v1', {})
         self.assertItemsEqual([], instance_data['base64_encoded_keys'])
-        self.assertEqual('nocloud', v1_data['cloud_name'])
+        self.assertEqual('unknown', v1_data['cloud_name'])
+        self.assertEqual('nocloud', v1_data['platform'])
+        self.assertEqual('config-disk (/dev/vda)', v1_data['subplatform'])
         self.assertIsNone(
             v1_data['availability_zone'],
             'found unexpected kvm availability_zone %s' %
diff --git a/tests/unittests/test_datasource/test_aliyun.py b/tests/unittests/test_datasource/test_aliyun.py
index 1e77842..e9213ca 100644
--- a/tests/unittests/test_datasource/test_aliyun.py
+++ b/tests/unittests/test_datasource/test_aliyun.py
@@ -140,6 +140,10 @@ class TestAliYunDatasource(test_helpers.HttprettyTestCase):
         self._test_get_sshkey()
         self._test_get_iid()
         self._test_host_name()
+        self.assertEqual('aliyun', self.ds.cloud_name)
+        self.assertEqual('ec2', self.ds.platform)
+        self.assertEqual(
+           'metadata (http://100.100.100.200)', self.ds.subplatform)
 
     @mock.patch("cloudinit.sources.DataSourceAliYun._is_aliyun")
     def test_returns_false_when_not_on_aliyun(self, m_is_aliyun):
diff --git a/tests/unittests/test_datasource/test_altcloud.py b/tests/unittests/test_datasource/test_altcloud.py
index ff35904..3119bfa 100644
--- a/tests/unittests/test_datasource/test_altcloud.py
+++ b/tests/unittests/test_datasource/test_altcloud.py
@@ -10,7 +10,6 @@
 This test file exercises the code in sources DataSourceAltCloud.py
 '''
 
-import mock
 import os
 import shutil
 import tempfile
@@ -18,32 +17,13 @@ import tempfile
 from cloudinit import helpers
 from cloudinit import util
 
-from cloudinit.tests.helpers import CiTestCase
+from cloudinit.tests.helpers import CiTestCase, mock
 
 import cloudinit.sources.DataSourceAltCloud as dsac
 
 OS_UNAME_ORIG = getattr(os, 'uname')
 
 
-def _write_cloud_info_file(value):
-    '''
-    Populate the CLOUD_INFO_FILE which would be populated
-    with a cloud backend identifier ImageFactory when building
-    an image with ImageFactory.
-    '''
-    cifile = open(dsac.CLOUD_INFO_FILE, 'w')
-    cifile.write(value)
-    cifile.close()
-    os.chmod(dsac.CLOUD_INFO_FILE, 0o664)
-
-
-def _remove_cloud_info_file():
-    '''
-    Remove the test CLOUD_INFO_FILE
-    '''
-    os.remove(dsac.CLOUD_INFO_FILE)
-
-
 def _write_user_data_files(mount_dir, value):
     '''
     Populate the deltacloud_user_data_file the user_data_file
@@ -98,13 +78,15 @@ def _dmi_data(expected):
 
 
 class TestGetCloudType(CiTestCase):
-    '''
-    Test to exercise method: DataSourceAltCloud.get_cloud_type()
-    '''
+    '''Test to exercise method: DataSourceAltCloud.get_cloud_type()'''
+
+    with_logs = True
 
     def setUp(self):
         '''Set up.'''
-        self.paths = helpers.Paths({'cloud_dir': '/tmp'})
+        super(TestGetCloudType, self).setUp()
+        self.tmp = self.tmp_dir()
+        self.paths = helpers.Paths({'cloud_dir': self.tmp})
         self.dmi_data = util.read_dmi_data
         # We have a different code path for arm to deal with LP1243287
         # We have to switch arch to x86_64 to avoid test failure
@@ -115,6 +97,26 @@ class TestGetCloudType(CiTestCase):
         util.read_dmi_data = self.dmi_data
         force_arch()
 
+    def test_cloud_info_file_ioerror(self):
+        """Return UNKNOWN when /etc/sysconfig/cloud-info exists but errors."""
+        self.assertEqual('/etc/sysconfig/cloud-info', dsac.CLOUD_INFO_FILE)
+        dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
+        # Attempting to read the directory generates IOError
+        with mock.patch.object(dsac, 'CLOUD_INFO_FILE', self.tmp):
+            self.assertEqual('UNKNOWN', dsrc.get_cloud_type())
+        self.assertIn(
+            "[Errno 21] Is a directory: '%s'" % self.tmp,
+            self.logs.getvalue())
+
+    def test_cloud_info_file(self):
+        """Return uppercase stripped content from /etc/sysconfig/cloud-info."""
+        dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
+        cloud_info = self.tmp_path('cloud-info', dir=self.tmp)
+        util.write_file(cloud_info, ' OverRiDdeN CloudType ')
+        # Attempting to read the directory generates IOError
+        with mock.patch.object(dsac, 'CLOUD_INFO_FILE', cloud_info):
+            self.assertEqual('OVERRIDDEN CLOUDTYPE', dsrc.get_cloud_type())
+
     def test_rhev(self):
         '''
         Test method get_cloud_type() for RHEVm systems.
@@ -153,60 +155,57 @@ class TestGetDataCloudInfoFile(CiTestCase):
         self.tmp = self.tmp_dir()
         self.paths = helpers.Paths(
             {'cloud_dir': self.tmp, 'run_dir': self.tmp})
-        self.cloud_info_file = tempfile.mkstemp()[1]
-        self.dmi_data = util.read_dmi_data
-        dsac.CLOUD_INFO_FILE = self.cloud_info_file
-
-    def tearDown(self):
-        # Reset
-
-        # Attempt to remove the temp file ignoring errors
-        try:
-            os.remove(self.cloud_info_file)
-        except OSError:
-            pass
-
-        util.read_dmi_data = self.dmi_data
-        dsac.CLOUD_INFO_FILE = '/etc/sysconfig/cloud-info'
+        self.cloud_info_file = self.tmp_path('cloud-info', dir=self.tmp)
 
     def test_rhev(self):
         '''Success Test module get_data() forcing RHEV.'''
 
-        _write_cloud_info_file('RHEV')
+        util.write_file(self.cloud_info_file, 'RHEV')
         dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
         dsrc.user_data_rhevm = lambda: True
-        self.assertEqual(True, dsrc.get_data())
+        with mock.patch.object(dsac, 'CLOUD_INFO_FILE', self.cloud_info_file):
+            self.assertEqual(True, dsrc.get_data())
+        self.assertEqual('altcloud', dsrc.cloud_name)
+        self.assertEqual('altcloud', dsrc.platform_type)
+        self.assertEqual('rhev (/dev/fd0)', dsrc.subplatform)
 
     def test_vsphere(self):
         '''Success Test module get_data() forcing VSPHERE.'''
 
-        _write_cloud_info_file('VSPHERE')
+        util.write_file(self.cloud_info_file, 'VSPHERE')
         dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
         dsrc.user_data_vsphere = lambda: True
-        self.assertEqual(True, dsrc.get_data())
+        with mock.patch.object(dsac, 'CLOUD_INFO_FILE', self.cloud_info_file):
+            self.assertEqual(True, dsrc.get_data())
+        self.assertEqual('altcloud', dsrc.cloud_name)
+        self.assertEqual('altcloud', dsrc.platform_type)
+        self.assertEqual('vsphere (unknown)', dsrc.subplatform)
 
     def test_fail_rhev(self):
         '''Failure Test module get_data() forcing RHEV.'''
 
-        _write_cloud_info_file('RHEV')
+        util.write_file(self.cloud_info_file, 'RHEV')
         dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
         dsrc.user_data_rhevm = lambda: False
-        self.assertEqual(False, dsrc.get_data())
+        with mock.patch.object(dsac, 'CLOUD_INFO_FILE', self.cloud_info_file):
+            self.assertEqual(False, dsrc.get_data())
 
     def test_fail_vsphere(self):
         '''Failure Test module get_data() forcing VSPHERE.'''
 
-        _write_cloud_info_file('VSPHERE')
+        util.write_file(self.cloud_info_file, 'VSPHERE')
         dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
         dsrc.user_data_vsphere = lambda: False
-        self.assertEqual(False, dsrc.get_data())
+        with mock.patch.object(dsac, 'CLOUD_INFO_FILE', self.cloud_info_file):
+            self.assertEqual(False, dsrc.get_data())
 
     def test_unrecognized(self):
         '''Failure Test module get_data() forcing unrecognized.'''
 
-        _write_cloud_info_file('unrecognized')
+        util.write_file(self.cloud_info_file, 'unrecognized')
         dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
-        self.assertEqual(False, dsrc.get_data())
+        with mock.patch.object(dsac, 'CLOUD_INFO_FILE', self.cloud_info_file):
+            self.assertEqual(False, dsrc.get_data())
 
 
 class TestGetDataNoCloudInfoFile(CiTestCase):
@@ -322,7 +321,8 @@ class TestUserDataVsphere(CiTestCase):
     '''
     def setUp(self):
         '''Set up.'''
-        self.paths = helpers.Paths({'cloud_dir': '/tmp'})
+        self.tmp = self.tmp_dir()
+        self.paths = helpers.Paths({'cloud_dir': self.tmp})
         self.mount_dir = tempfile.mkdtemp()
 
         _write_user_data_files(self.mount_dir, 'test user data')
@@ -363,6 +363,22 @@ class TestUserDataVsphere(CiTestCase):
         self.assertEqual(1, m_find_devs_with.call_count)
         self.assertEqual(1, m_mount_cb.call_count)
 
+    @mock.patch("cloudinit.sources.DataSourceAltCloud.util.find_devs_with")
+    @mock.patch("cloudinit.sources.DataSourceAltCloud.util.mount_cb")
+    def test_user_data_vsphere_success(self, m_mount_cb, m_find_devs_with):
+        """Test user_data_vsphere() where successful."""
+        m_find_devs_with.return_value = ["/dev/mock/cdrom"]
+        m_mount_cb.return_value = 'raw userdata from cdrom'
+        dsrc = dsac.DataSourceAltCloud({}, None, self.paths)
+        cloud_info = self.tmp_path('cloud-info', dir=self.tmp)
+        util.write_file(cloud_info, 'VSPHERE')
+        self.assertEqual(True, dsrc.user_data_vsphere())
+        m_find_devs_with.assert_called_once_with('LABEL=CDROM')
+        m_mount_cb.assert_called_once_with(
+            '/dev/mock/cdrom', dsac.read_user_data_callback)
+        with mock.patch.object(dsrc, 'get_cloud_type', return_value='VSPHERE'):
+            self.assertEqual('vsphere (/dev/mock/cdrom)', dsrc.subplatform)
+
 
 class TestReadUserDataCallback(CiTestCase):
     '''
diff --git a/tests/unittests/test_datasource/test_azure.py b/tests/unittests/test_datasource/test_azure.py
index 4e428b7..0f4b7bf 100644
--- a/tests/unittests/test_datasource/test_azure.py
+++ b/tests/unittests/test_datasource/test_azure.py
@@ -110,6 +110,8 @@ NETWORK_METADATA = {
     }
 }
 
+MOCKPATH = 'cloudinit.sources.DataSourceAzure.'
+
 
 class TestGetMetadataFromIMDS(HttprettyTestCase):
 
@@ -119,9 +121,9 @@ class TestGetMetadataFromIMDS(HttprettyTestCase):
         super(TestGetMetadataFromIMDS, self).setUp()
         self.network_md_url = dsaz.IMDS_URL + "instance?api-version=2017-12-01"
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.readurl')
-    @mock.patch('cloudinit.sources.DataSourceAzure.EphemeralDHCPv4')
-    @mock.patch('cloudinit.sources.DataSourceAzure.net.is_up')
+    @mock.patch(MOCKPATH + 'readurl')
+    @mock.patch(MOCKPATH + 'EphemeralDHCPv4')
+    @mock.patch(MOCKPATH + 'net.is_up')
     def test_get_metadata_does_not_dhcp_if_network_is_up(
             self, m_net_is_up, m_dhcp, m_readurl):
         """Do not perform DHCP setup when nic is already up."""
@@ -138,9 +140,9 @@ class TestGetMetadataFromIMDS(HttprettyTestCase):
             "Crawl of Azure Instance Metadata Service (IMDS) took",  # log_time
             self.logs.getvalue())
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.readurl')
-    @mock.patch('cloudinit.sources.DataSourceAzure.EphemeralDHCPv4')
-    @mock.patch('cloudinit.sources.DataSourceAzure.net.is_up')
+    @mock.patch(MOCKPATH + 'readurl')
+    @mock.patch(MOCKPATH + 'EphemeralDHCPv4')
+    @mock.patch(MOCKPATH + 'net.is_up')
     def test_get_metadata_performs_dhcp_when_network_is_down(
             self, m_net_is_up, m_dhcp, m_readurl):
         """Perform DHCP setup when nic is not up."""
@@ -163,7 +165,7 @@ class TestGetMetadataFromIMDS(HttprettyTestCase):
             headers={'Metadata': 'true'}, retries=2, timeout=1)
 
     @mock.patch('cloudinit.url_helper.time.sleep')
-    @mock.patch('cloudinit.sources.DataSourceAzure.net.is_up')
+    @mock.patch(MOCKPATH + 'net.is_up')
     def test_get_metadata_from_imds_empty_when_no_imds_present(
             self, m_net_is_up, m_sleep):
         """Return empty dict when IMDS network metadata is absent."""
@@ -380,7 +382,7 @@ fdescfs            /dev/fd          fdescfs rw              0 0
             res = get_path_dev_freebsd('/etc', mnt_list)
             self.assertIsNotNone(res)
 
-    @mock.patch('cloudinit.sources.DataSourceAzure._is_platform_viable')
+    @mock.patch(MOCKPATH + '_is_platform_viable')
     def test_call_is_platform_viable_seed(self, m_is_platform_viable):
         """Check seed_dir using _is_platform_viable and return False."""
         # Return a non-matching asset tag value
@@ -401,6 +403,24 @@ fdescfs            /dev/fd          fdescfs rw              0 0
         self.assertEqual(dsrc.metadata['local-hostname'], odata['HostName'])
         self.assertTrue(os.path.isfile(
             os.path.join(self.waagent_d, 'ovf-env.xml')))
+        self.assertEqual('azure', dsrc.cloud_name)
+        self.assertEqual('azure', dsrc.platform_type)
+        self.assertEqual(
+            'seed-dir (%s/seed/azure)' % self.tmp, dsrc.subplatform)
+
+    def test_basic_dev_file(self):
+        """When a device path is used, present that in subplatform."""
+        data = {'sys_cfg': {}, 'dsdevs': ['/dev/cd0']}
+        dsrc = self._get_ds(data)
+        with mock.patch(MOCKPATH + 'util.mount_cb') as m_mount_cb:
+            m_mount_cb.return_value = (
+                {'local-hostname': 'me'}, 'ud', {'cfg': ''}, {})
+            self.assertTrue(dsrc.get_data())
+        self.assertEqual(dsrc.userdata_raw, 'ud')
+        self.assertEqual(dsrc.metadata['local-hostname'], 'me')
+        self.assertEqual('azure', dsrc.cloud_name)
+        self.assertEqual('azure', dsrc.platform_type)
+        self.assertEqual('config-disk (/dev/cd0)', dsrc.subplatform)
 
     def test_get_data_non_ubuntu_will_not_remove_network_scripts(self):
         """get_data on non-Ubuntu will not remove ubuntu net scripts."""
@@ -769,8 +789,8 @@ fdescfs            /dev/fd          fdescfs rw              0 0
         ds.get_data()
         self.assertEqual(self.instance_id, ds.metadata['instance-id'])
 
-    @mock.patch("cloudinit.sources.DataSourceAzure.util.is_FreeBSD")
-    @mock.patch("cloudinit.sources.DataSourceAzure._check_freebsd_cdrom")
+    @mock.patch(MOCKPATH + 'util.is_FreeBSD')
+    @mock.patch(MOCKPATH + '_check_freebsd_cdrom')
     def test_list_possible_azure_ds_devs(self, m_check_fbsd_cdrom,
                                          m_is_FreeBSD):
         """On FreeBSD, possible devs should show /dev/cd0."""
@@ -885,17 +905,17 @@ fdescfs            /dev/fd          fdescfs rw              0 0
         expected_config['config'].append(blacklist_config)
         self.assertEqual(netconfig, expected_config)
 
-    @mock.patch("cloudinit.sources.DataSourceAzure.util.subp")
+    @mock.patch(MOCKPATH + 'util.subp')
     def test_get_hostname_with_no_args(self, subp):
         dsaz.get_hostname()
         subp.assert_called_once_with(("hostname",), capture=True)
 
-    @mock.patch("cloudinit.sources.DataSourceAzure.util.subp")
+    @mock.patch(MOCKPATH + 'util.subp')
     def test_get_hostname_with_string_arg(self, subp):
         dsaz.get_hostname(hostname_command="hostname")
         subp.assert_called_once_with(("hostname",), capture=True)
 
-    @mock.patch("cloudinit.sources.DataSourceAzure.util.subp")
+    @mock.patch(MOCKPATH + 'util.subp')
     def test_get_hostname_with_iterable_arg(self, subp):
         dsaz.get_hostname(hostname_command=("hostname",))
         subp.assert_called_once_with(("hostname",), capture=True)
@@ -949,7 +969,7 @@ class TestAzureBounce(CiTestCase):
         self.set_hostname = self.patches.enter_context(
             mock.patch.object(dsaz, 'set_hostname'))
         self.subp = self.patches.enter_context(
-            mock.patch('cloudinit.sources.DataSourceAzure.util.subp'))
+            mock.patch(MOCKPATH + 'util.subp'))
         self.find_fallback_nic = self.patches.enter_context(
             mock.patch('cloudinit.net.find_fallback_nic', return_value='eth9'))
 
@@ -989,7 +1009,7 @@ class TestAzureBounce(CiTestCase):
         ds.get_data()
         self.assertEqual(0, self.set_hostname.call_count)
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.perform_hostname_bounce')
+    @mock.patch(MOCKPATH + 'perform_hostname_bounce')
     def test_disabled_bounce_does_not_perform_bounce(
             self, perform_hostname_bounce):
         cfg = {'hostname_bounce': {'policy': 'off'}}
@@ -1005,7 +1025,7 @@ class TestAzureBounce(CiTestCase):
         ds.get_data()
         self.assertEqual(0, self.set_hostname.call_count)
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.perform_hostname_bounce')
+    @mock.patch(MOCKPATH + 'perform_hostname_bounce')
     def test_unchanged_hostname_does_not_perform_bounce(
             self, perform_hostname_bounce):
         host_name = 'unchanged-host-name'
@@ -1015,7 +1035,7 @@ class TestAzureBounce(CiTestCase):
         ds.get_data()
         self.assertEqual(0, perform_hostname_bounce.call_count)
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.perform_hostname_bounce')
+    @mock.patch(MOCKPATH + 'perform_hostname_bounce')
     def test_force_performs_bounce_regardless(self, perform_hostname_bounce):
         host_name = 'unchanged-host-name'
         self.get_hostname.return_value = host_name
@@ -1032,7 +1052,7 @@ class TestAzureBounce(CiTestCase):
         cfg = {'hostname_bounce': {'policy': 'force'}}
         dsrc = self._get_ds(self.get_ovf_env_with_dscfg(host_name, cfg),
                             agent_command=['not', '__builtin__'])
-        patch_path = 'cloudinit.sources.DataSourceAzure.util.which'
+        patch_path = MOCKPATH + 'util.which'
         with mock.patch(patch_path) as m_which:
             m_which.return_value = None
             ret = self._get_and_setup(dsrc)
@@ -1053,7 +1073,7 @@ class TestAzureBounce(CiTestCase):
         self.assertEqual(expected_hostname,
                          self.set_hostname.call_args_list[0][0][0])
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.perform_hostname_bounce')
+    @mock.patch(MOCKPATH + 'perform_hostname_bounce')
     def test_different_hostnames_performs_bounce(
             self, perform_hostname_bounce):
         expected_hostname = 'azure-expected-host-name'
@@ -1076,7 +1096,7 @@ class TestAzureBounce(CiTestCase):
         self.assertEqual(initial_host_name,
                          self.set_hostname.call_args_list[-1][0][0])
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.perform_hostname_bounce')
+    @mock.patch(MOCKPATH + 'perform_hostname_bounce')
     def test_failure_in_bounce_still_resets_host_name(
             self, perform_hostname_bounce):
         perform_hostname_bounce.side_effect = Exception
@@ -1117,7 +1137,7 @@ class TestAzureBounce(CiTestCase):
         self.assertEqual(
             dsaz.BOUNCE_COMMAND_IFUP, bounce_args)
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.perform_hostname_bounce')
+    @mock.patch(MOCKPATH + 'perform_hostname_bounce')
     def test_set_hostname_option_can_disable_bounce(
             self, perform_hostname_bounce):
         cfg = {'set_hostname': False, 'hostname_bounce': {'policy': 'force'}}
@@ -1218,12 +1238,12 @@ class TestCanDevBeReformatted(CiTestCase):
         def has_ntfs_fs(device):
             return bypath.get(device, {}).get('fs') == 'ntfs'
 
-        p = 'cloudinit.sources.DataSourceAzure'
-        self._domock(p + "._partitions_on_device", 'm_partitions_on_device')
-        self._domock(p + "._has_ntfs_filesystem", 'm_has_ntfs_filesystem')
-        self._domock(p + ".util.mount_cb", 'm_mount_cb')
-        self._domock(p + ".os.path.realpath", 'm_realpath')
-        self._domock(p + ".os.path.exists", 'm_exists')
+        p = MOCKPATH
+        self._domock(p + "_partitions_on_device", 'm_partitions_on_device')
+        self._domock(p + "_has_ntfs_filesystem", 'm_has_ntfs_filesystem')
+        self._domock(p + "util.mount_cb", 'm_mount_cb')
+        self._domock(p + "os.path.realpath", 'm_realpath')
+        self._domock(p + "os.path.exists", 'm_exists')
 
         self.m_exists.side_effect = lambda p: p in bypath
         self.m_realpath.side_effect = realpath
@@ -1488,7 +1508,7 @@ class TestPreprovisioningShouldReprovision(CiTestCase):
         self.paths = helpers.Paths({'cloud_dir': tmp})
         dsaz.BUILTIN_DS_CONFIG['data_dir'] = self.waagent_d
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.util.write_file')
+    @mock.patch(MOCKPATH + 'util.write_file')
     def test__should_reprovision_with_true_cfg(self, isfile, write_f):
         """The _should_reprovision method should return true with config
            flag present."""
@@ -1512,7 +1532,7 @@ class TestPreprovisioningShouldReprovision(CiTestCase):
         dsa = dsaz.DataSourceAzure({}, distro=None, paths=self.paths)
         self.assertFalse(dsa._should_reprovision((None, None, {}, None)))
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.DataSourceAzure._poll_imds')
+    @mock.patch(MOCKPATH + 'DataSourceAzure._poll_imds')
     def test_reprovision_calls__poll_imds(self, _poll_imds, isfile):
         """_reprovision will poll IMDS."""
         isfile.return_value = False
@@ -1528,8 +1548,7 @@ class TestPreprovisioningShouldReprovision(CiTestCase):
 @mock.patch('cloudinit.net.dhcp.EphemeralIPv4Network')
 @mock.patch('cloudinit.net.dhcp.maybe_perform_dhcp_discovery')
 @mock.patch('requests.Session.request')
[e-mail](
-    'cloudinit.sources.DataSourceAzure.DataSourceAzure._report_ready')
[e-mail](MOCKPATH + 'DataSourceAzure._report_ready')
 class TestPreprovisioningPollIMDS(CiTestCase):
 
     def setUp(self):
@@ -1539,7 +1558,7 @@ class TestPreprovisioningPollIMDS(CiTestCase):
         self.paths = helpers.Paths({'cloud_dir': self.tmp})
         dsaz.BUILTIN_DS_CONFIG['data_dir'] = self.waagent_d
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.util.write_file')
+    @mock.patch(MOCKPATH + 'util.write_file')
     def test_poll_imds_calls_report_ready(self, write_f, report_ready_func,
                                           fake_resp, m_dhcp, m_net):
         """The poll_imds will call report_ready after creating marker file."""
@@ -1550,8 +1569,7 @@ class TestPreprovisioningPollIMDS(CiTestCase):
             'unknown-245': '624c3620'}
         m_dhcp.return_value = [lease]
         dsa = dsaz.DataSourceAzure({}, distro=None, paths=self.paths)
-        mock_path = (
-            'cloudinit.sources.DataSourceAzure.REPORTED_READY_MARKER_FILE')
+        mock_path = (MOCKPATH + 'REPORTED_READY_MARKER_FILE')
         with mock.patch(mock_path, report_marker):
             dsa._poll_imds()
         self.assertEqual(report_ready_func.call_count, 1)
@@ -1561,23 +1579,21 @@ class TestPreprovisioningPollIMDS(CiTestCase):
                                           fake_resp, m_dhcp, m_net):
         """The poll_imds should not call reporting ready
            when flag is false"""
-        report_marker = self.tmp_path('report_marker', self.tmp)
-        write_file(report_marker, content='dont run report_ready :)')
+        report_file = self.tmp_path('report_marker', self.tmp)
+        write_file(report_file, content='dont run report_ready :)')
         m_dhcp.return_value = [{
             'interface': 'eth9', 'fixed-address': '192.168.2.9',
             'routers': '192.168.2.1', 'subnet-mask': '255.255.255.0',
             'unknown-245': '624c3620'}]
         dsa = dsaz.DataSourceAzure({}, distro=None, paths=self.paths)
-        mock_path = (
-            'cloudinit.sources.DataSourceAzure.REPORTED_READY_MARKER_FILE')
-        with mock.patch(mock_path, report_marker):
+        with mock.patch(MOCKPATH + 'REPORTED_READY_MARKER_FILE', report_file):
             dsa._poll_imds()
         self.assertEqual(report_ready_func.call_count, 0)
 
 
[e-mail]('cloudinit.sources.DataSourceAzure.util.subp')
[e-mail]('cloudinit.sources.DataSourceAzure.util.write_file')
[e-mail]('cloudinit.sources.DataSourceAzure.util.is_FreeBSD')
[e-mail](MOCKPATH + 'util.subp')
[e-mail](MOCKPATH + 'util.write_file')
[e-mail](MOCKPATH + 'util.is_FreeBSD')
 @mock.patch('cloudinit.net.dhcp.EphemeralIPv4Network')
 @mock.patch('cloudinit.net.dhcp.maybe_perform_dhcp_discovery')
 @mock.patch('requests.Session.request')
@@ -1688,7 +1704,7 @@ class TestRemoveUbuntuNetworkConfigScripts(CiTestCase):
             self.tmp_path('notfilehere', dir=self.tmp)])
         self.assertNotIn('/not/a', self.logs.getvalue())  # No delete logs
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.os.path.exists')
+    @mock.patch(MOCKPATH + 'os.path.exists')
     def test_remove_network_scripts_default_removes_stock_scripts(self,
                                                                   m_exists):
         """Azure's stock ubuntu image scripts and artifacts are removed."""
@@ -1704,14 +1720,14 @@ class TestWBIsPlatformViable(CiTestCase):
     """White box tests for _is_platform_viable."""
     with_logs = True
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.util.read_dmi_data')
+    @mock.patch(MOCKPATH + 'util.read_dmi_data')
     def test_true_on_non_azure_chassis(self, m_read_dmi_data):
         """Return True if DMI chassis-asset-tag is AZURE_CHASSIS_ASSET_TAG."""
         m_read_dmi_data.return_value = dsaz.AZURE_CHASSIS_ASSET_TAG
         self.assertTrue(dsaz._is_platform_viable('doesnotmatter'))
 
-    @mock.patch('cloudinit.sources.DataSourceAzure.os.path.exists')
-    @mock.patch('cloudinit.sources.DataSourceAzure.util.read_dmi_data')
+    @mock.patch(MOCKPATH + 'os.path.exists')
+    @mock.patch(MOCKPATH + 'util.read_dmi_data')
     def test_true_on_azure_ovf_env_in_seed_dir(self, m_read_dmi_data, m_exist):
         """Return True if ovf-env.xml exists in known seed dirs."""
         # Non-matching Azure chassis-asset-tag
@@ -1729,7 +1745,7 @@ class TestWBIsPlatformViable(CiTestCase):
         and no devices have a label starting with prefix 'rd_rdfe_'.
         """
         self.assertFalse(wrap_and_call(
-            'cloudinit.sources.DataSourceAzure',
+            MOCKPATH,
             {'os.path.exists': False,
              # Non-matching Azure chassis-asset-tag
              'util.read_dmi_data': dsaz.AZURE_CHASSIS_ASSET_TAG + 'X',
diff --git a/tests/unittests/test_datasource/test_cloudsigma.py b/tests/unittests/test_datasource/test_cloudsigma.py
index 380ad1b..3bf52e6 100644
--- a/tests/unittests/test_datasource/test_cloudsigma.py
+++ b/tests/unittests/test_datasource/test_cloudsigma.py
@@ -68,6 +68,12 @@ class DataSourceCloudSigmaTest(test_helpers.CiTestCase):
         self.assertEqual(SERVER_CONTEXT['uuid'],
                          self.datasource.get_instance_id())
 
+    def test_platform(self):
+        """All platform-related attributes are set."""
+        self.assertEqual(self.datasource.cloud_name, 'cloudsigma')
+        self.assertEqual(self.datasource.platform_type, 'cloudsigma')
+        self.assertEqual(self.datasource.subplatform, 'cepko (/dev/ttyS1)')
+
     def test_metadata(self):
         self.assertEqual(self.datasource.metadata, SERVER_CONTEXT)
 
diff --git a/tests/unittests/test_datasource/test_configdrive.py b/tests/unittests/test_datasource/test_configdrive.py
index 231619c..dcdabea 100644
--- a/tests/unittests/test_datasource/test_configdrive.py
+++ b/tests/unittests/test_datasource/test_configdrive.py
@@ -478,6 +478,9 @@ class TestConfigDriveDataSource(CiTestCase):
         myds = cfg_ds_from_dir(self.tmp, files=CFG_DRIVE_FILES_V2)
         self.assertEqual(myds.get_public_ssh_keys(),
                          [OSTACK_META['public_keys']['mykey']])
+        self.assertEqual('configdrive', myds.cloud_name)
+        self.assertEqual('openstack', myds.platform)
+        self.assertEqual('seed-dir (%s/seed)' % self.tmp, myds.subplatform)
 
 
 class TestNetJson(CiTestCase):
diff --git a/tests/unittests/test_datasource/test_ec2.py b/tests/unittests/test_datasource/test_ec2.py
index 497e761..9f81255 100644
--- a/tests/unittests/test_datasource/test_ec2.py
+++ b/tests/unittests/test_datasource/test_ec2.py
@@ -351,7 +351,9 @@ class TestEc2(test_helpers.HttprettyTestCase):
             m_get_interface_mac.return_value = mac1
             nc = ds.network_config  # Will re-crawl network metadata
             self.assertIsNotNone(nc)
-        self.assertIn('Re-crawl of metadata service', self.logs.getvalue())
+        self.assertIn(
+            'Refreshing stale metadata from prior to upgrade',
+            self.logs.getvalue())
         expected = {'version': 1, 'config': [
             {'mac_address': '06:17:04:d7:26:09',
              'name': 'eth9',
@@ -386,7 +388,7 @@ class TestEc2(test_helpers.HttprettyTestCase):
         register_mock_metaserver(
             '{0}/{1}/dynamic/'.format(ds.metadata_address, all_versions[-1]),
             DYNAMIC_METADATA)
-        ds._cloud_platform = ec2.Platforms.AWS
+        ds._cloud_name = ec2.CloudNames.AWS
         # Setup cached metadata on the Datasource
         ds.metadata = DEFAULT_METADATA
         self.assertEqual('my-identity-id', ds.get_instance_id())
@@ -401,6 +403,9 @@ class TestEc2(test_helpers.HttprettyTestCase):
         ret = ds.get_data()
         self.assertTrue(ret)
         self.assertEqual(0, m_dhcp.call_count)
+        self.assertEqual('aws', ds.cloud_name)
+        self.assertEqual('ec2', ds.platform_type)
+        self.assertEqual('metadata (%s)' % ds.metadata_address, ds.subplatform)
 
     def test_valid_platform_with_strict_false(self):
         """Valid platform data should return true with strict_id false."""
@@ -439,16 +444,17 @@ class TestEc2(test_helpers.HttprettyTestCase):
             sys_cfg={'datasource': {'Ec2': {'strict_id': False}}},
             md=DEFAULT_METADATA)
         platform_attrs = [
-            attr for attr in ec2.Platforms.__dict__.keys()
+            attr for attr in ec2.CloudNames.__dict__.keys()
             if not attr.startswith('__')]
         for attr_name in platform_attrs:
-            platform_name = getattr(ec2.Platforms, attr_name)
-            if platform_name != 'AWS':
-                ds._cloud_platform = platform_name
+            platform_name = getattr(ec2.CloudNames, attr_name)
+            if platform_name != 'aws':
+                ds._cloud_name = platform_name
                 ret = ds.get_data()
+                self.assertEqual('ec2', ds.platform_type)
                 self.assertFalse(ret)
                 message = (
-                    "Local Ec2 mode only supported on ('AWS',),"
+                    "Local Ec2 mode only supported on ('aws',),"
                     ' not {0}'.format(platform_name))
                 self.assertIn(message, self.logs.getvalue())
 
diff --git a/tests/unittests/test_datasource/test_ibmcloud.py b/tests/unittests/test_datasource/test_ibmcloud.py
index e639ae4..0b54f58 100644
--- a/tests/unittests/test_datasource/test_ibmcloud.py
+++ b/tests/unittests/test_datasource/test_ibmcloud.py
@@ -1,14 +1,17 @@
 # This file is part of cloud-init. See LICENSE file for license information.
 
+from cloudinit.helpers import Paths
 from cloudinit.sources import DataSourceIBMCloud as ibm
 from cloudinit.tests import helpers as test_helpers
+from cloudinit import util
 
 import base64
 import copy
 import json
-import mock
 from textwrap import dedent
 
+mock = test_helpers.mock
+
 D_PATH = "cloudinit.sources.DataSourceIBMCloud."
 
 
@@ -309,4 +312,39 @@ class TestIsIBMProvisioning(test_helpers.FilesystemMockingTestCase):
         self.assertIn("no reference file", self.logs.getvalue())
 
 
+class TestDataSourceIBMCloud(test_helpers.CiTestCase):
+
+    def setUp(self):
+        super(TestDataSourceIBMCloud, self).setUp()
+        self.tmp = self.tmp_dir()
+        self.cloud_dir = self.tmp_path('cloud', dir=self.tmp)
+        util.ensure_dir(self.cloud_dir)
+        paths = Paths({'run_dir': self.tmp, 'cloud_dir': self.cloud_dir})
+        self.ds = ibm.DataSourceIBMCloud(
+            sys_cfg={}, distro=None, paths=paths)
+
+    def test_get_data_false(self):
+        """When read_md returns None, get_data returns False."""
+        with mock.patch(D_PATH + 'read_md', return_value=None):
+            self.assertFalse(self.ds.get_data())
+
+    def test_get_data_processes_read_md(self):
+        """get_data processes and caches content returned by read_md."""
+        md = {
+            'metadata': {}, 'networkdata': 'net', 'platform': 'plat',
+            'source': 'src', 'system-uuid': 'uuid', 'userdata': 'ud',
+            'vendordata': 'vd'}
+        with mock.patch(D_PATH + 'read_md', return_value=md):
+            self.assertTrue(self.ds.get_data())
+        self.assertEqual('src', self.ds.source)
+        self.assertEqual('plat', self.ds.platform)
+        self.assertEqual({}, self.ds.metadata)
+        self.assertEqual('ud', self.ds.userdata_raw)
+        self.assertEqual('net', self.ds.network_json)
+        self.assertEqual('vd', self.ds.vendordata_pure)
+        self.assertEqual('uuid', self.ds.system_uuid)
+        self.assertEqual('ibmcloud', self.ds.cloud_name)
+        self.assertEqual('ibmcloud', self.ds.platform_type)
+        self.assertEqual('plat (src)', self.ds.subplatform)
+
 # vi: ts=4 expandtab
diff --git a/tests/unittests/test_datasource/test_nocloud.py b/tests/unittests/test_datasource/test_nocloud.py
index 21931eb..b6468b6 100644
--- a/tests/unittests/test_datasource/test_nocloud.py
+++ b/tests/unittests/test_datasource/test_nocloud.py
@@ -10,6 +10,7 @@ import textwrap
 import yaml
 
 
[e-mail]('cloudinit.sources.DataSourceNoCloud.util.is_lxd')
 class TestNoCloudDataSource(CiTestCase):
 
     def setUp(self):
@@ -28,10 +29,11 @@ class TestNoCloudDataSource(CiTestCase):
         self.mocks.enter_context(
             mock.patch.object(util, 'read_dmi_data', return_value=None))
 
-    def test_nocloud_seed_dir(self):
+    def test_nocloud_seed_dir_on_lxd(self, m_is_lxd):
         md = {'instance-id': 'IID', 'dsmode': 'local'}
         ud = b"USER_DATA_HERE"
-        populate_dir(os.path.join(self.paths.seed_dir, "nocloud"),
+        seed_dir = os.path.join(self.paths.seed_dir, "nocloud")
+        populate_dir(seed_dir,
                      {'user-data': ud, 'meta-data': yaml.safe_dump(md)})
 
         sys_cfg = {
@@ -44,9 +46,32 @@ class TestNoCloudDataSource(CiTestCase):
         ret = dsrc.get_data()
         self.assertEqual(dsrc.userdata_raw, ud)
         self.assertEqual(dsrc.metadata, md)
+        self.assertEqual(dsrc.platform_type, 'lxd')
+        self.assertEqual(
+            dsrc.subplatform, 'seed-dir (%s)' % seed_dir)
         self.assertTrue(ret)
 
-    def test_fs_label(self):
+    def test_nocloud_seed_dir_non_lxd_platform_is_nocloud(self, m_is_lxd):
+        """Non-lxd environments will list nocloud as the platform."""
+        m_is_lxd.return_value = False
+        md = {'instance-id': 'IID', 'dsmode': 'local'}
+        seed_dir = os.path.join(self.paths.seed_dir, "nocloud")
+        populate_dir(seed_dir,
+                     {'user-data': '', 'meta-data': yaml.safe_dump(md)})
+
+        sys_cfg = {
+            'datasource': {'NoCloud': {'fs_label': None}}
+        }
+
+        ds = DataSourceNoCloud.DataSourceNoCloud
+
+        dsrc = ds(sys_cfg=sys_cfg, distro=None, paths=self.paths)
+        self.assertTrue(dsrc.get_data())
+        self.assertEqual(dsrc.platform_type, 'nocloud')
+        self.assertEqual(
+            dsrc.subplatform, 'seed-dir (%s)' % seed_dir)
+
+    def test_fs_label(self, m_is_lxd):
         # find_devs_with should not be called ff fs_label is None
         ds = DataSourceNoCloud.DataSourceNoCloud
 
@@ -68,7 +93,7 @@ class TestNoCloudDataSource(CiTestCase):
         ret = dsrc.get_data()
         self.assertFalse(ret)
 
-    def test_no_datasource_expected(self):
+    def test_no_datasource_expected(self, m_is_lxd):
         # no source should be found if no cmdline, config, and fs_label=None
         sys_cfg = {'datasource': {'NoCloud': {'fs_label': None}}}
 
@@ -76,7 +101,7 @@ class TestNoCloudDataSource(CiTestCase):
         dsrc = ds(sys_cfg=sys_cfg, distro=None, paths=self.paths)
         self.assertFalse(dsrc.get_data())
 
-    def test_seed_in_config(self):
+    def test_seed_in_config(self, m_is_lxd):
         ds = DataSourceNoCloud.DataSourceNoCloud
 
         data = {
@@ -92,7 +117,7 @@ class TestNoCloudDataSource(CiTestCase):
         self.assertEqual(dsrc.metadata.get('instance-id'), 'IID')
         self.assertTrue(ret)
 
-    def test_nocloud_seed_with_vendordata(self):
+    def test_nocloud_seed_with_vendordata(self, m_is_lxd):
         md = {'instance-id': 'IID', 'dsmode': 'local'}
         ud = b"USER_DATA_HERE"
         vd = b"THIS IS MY VENDOR_DATA"
@@ -114,7 +139,7 @@ class TestNoCloudDataSource(CiTestCase):
         self.assertEqual(dsrc.vendordata_raw, vd)
         self.assertTrue(ret)
 
-    def test_nocloud_no_vendordata(self):
+    def test_nocloud_no_vendordata(self, m_is_lxd):
         populate_dir(os.path.join(self.paths.seed_dir, "nocloud"),
                      {'user-data': b"ud", 'meta-data': "instance-id: IID\n"})
 
@@ -128,7 +153,7 @@ class TestNoCloudDataSource(CiTestCase):
         self.assertFalse(dsrc.vendordata)
         self.assertTrue(ret)
 
-    def test_metadata_network_interfaces(self):
+    def test_metadata_network_interfaces(self, m_is_lxd):
         gateway = "103.225.10.1"
         md = {
             'instance-id': 'i-abcd',
@@ -157,7 +182,7 @@ class TestNoCloudDataSource(CiTestCase):
         # very simple check just for the strings above
         self.assertIn(gateway, str(dsrc.network_config))
 
-    def test_metadata_network_config(self):
+    def test_metadata_network_config(self, m_is_lxd):
         # network-config needs to get into network_config
         netconf = {'version': 1,
                    'config': [{'type': 'physical', 'name': 'interface0',
@@ -177,7 +202,7 @@ class TestNoCloudDataSource(CiTestCase):
         self.assertTrue(ret)
         self.assertEqual(netconf, dsrc.network_config)
 
-    def test_metadata_network_config_over_interfaces(self):
+    def test_metadata_network_config_over_interfaces(self, m_is_lxd):
         # network-config should override meta-data/network-interfaces
         gateway = "103.225.10.1"
         md = {
diff --git a/tests/unittests/test_datasource/test_opennebula.py b/tests/unittests/test_datasource/test_opennebula.py
index 6159101..bb399f6 100644
--- a/tests/unittests/test_datasource/test_opennebula.py
+++ b/tests/unittests/test_datasource/test_opennebula.py
@@ -123,6 +123,10 @@ class TestOpenNebulaDataSource(CiTestCase):
             self.assertTrue(ret)
         finally:
             util.find_devs_with = orig_find_devs_with
+        self.assertEqual('opennebula', dsrc.cloud_name)
+        self.assertEqual('opennebula', dsrc.platform_type)
+        self.assertEqual(
+            'seed-dir (%s/seed/opennebula)' % self.tmp, dsrc.subplatform)
 
     def test_seed_dir_non_contextdisk(self):
         self.assertRaises(ds.NonContextDiskDir, ds.read_context_disk_dir,
diff --git a/tests/unittests/test_datasource/test_ovf.py b/tests/unittests/test_datasource/test_ovf.py
index 9d52eb9..a226c03 100644
--- a/tests/unittests/test_datasource/test_ovf.py
+++ b/tests/unittests/test_datasource/test_ovf.py
@@ -11,7 +11,7 @@ from collections import OrderedDict
 from textwrap import dedent
 
 from cloudinit import util
-from cloudinit.tests.helpers import CiTestCase, wrap_and_call
+from cloudinit.tests.helpers import CiTestCase, mock, wrap_and_call
 from cloudinit.helpers import Paths
 from cloudinit.sources import DataSourceOVF as dsovf
 from cloudinit.sources.helpers.vmware.imc.config_custom_script import (
@@ -120,7 +120,7 @@ class TestDatasourceOVF(CiTestCase):
 
     def test_get_data_false_on_none_dmi_data(self):
         """When dmi for system-product-name is None, get_data returns False."""
-        paths = Paths({'seed_dir': self.tdir})
+        paths = Paths({'cloud_dir': self.tdir})
         ds = self.datasource(sys_cfg={}, distro={}, paths=paths)
         retcode = wrap_and_call(
             'cloudinit.sources.DataSourceOVF',
@@ -134,7 +134,7 @@ class TestDatasourceOVF(CiTestCase):
 
     def test_get_data_no_vmware_customization_disabled(self):
         """When vmware customization is disabled via sys_cfg log a message."""
-        paths = Paths({'seed_dir': self.tdir})
+        paths = Paths({'cloud_dir': self.tdir})
         ds = self.datasource(
             sys_cfg={'disable_vmware_customization': True}, distro={},
             paths=paths)
@@ -153,7 +153,7 @@ class TestDatasourceOVF(CiTestCase):
         """When cloud-init workflow for vmware is enabled via sys_cfg log a
         message.
         """
-        paths = Paths({'seed_dir': self.tdir})
+        paths = Paths({'cloud_dir': self.tdir})
         ds = self.datasource(
             sys_cfg={'disable_vmware_customization': False}, distro={},
             paths=paths)
@@ -178,6 +178,50 @@ class TestDatasourceOVF(CiTestCase):
         self.assertIn('Script %s not found!!' % customscript,
                       str(context.exception))
 
+    def test_get_data_non_vmware_seed_platform_info(self):
+        """Platform info properly reports when on non-vmware platforms."""
+        paths = Paths({'cloud_dir': self.tdir, 'run_dir': self.tdir})
+        # Write ovf-env.xml seed file
+        seed_dir = self.tmp_path('seed', dir=self.tdir)
+        ovf_env = self.tmp_path('ovf-env.xml', dir=seed_dir)
+        util.write_file(ovf_env, OVF_ENV_CONTENT)
+        ds = self.datasource(sys_cfg={}, distro={}, paths=paths)
+
+        self.assertEqual('ovf', ds.cloud_name)
+        self.assertEqual('ovf', ds.platform_type)
+        MPATH = 'cloudinit.sources.DataSourceOVF.'
+        with mock.patch(MPATH + 'util.read_dmi_data', return_value='!VMware'):
+            with mock.patch(MPATH + 'transport_vmware_guestd') as m_guestd:
+                with mock.patch(MPATH + 'transport_iso9660') as m_iso9660:
+                    m_iso9660.return_value = (None, 'ignored', 'ignored')
+                    m_guestd.return_value = (None, 'ignored', 'ignored')
+                    self.assertTrue(ds.get_data())
+                    self.assertEqual(
+                        'ovf (%s/seed/ovf-env.xml)' % self.tdir,
+                        ds.subplatform)
+
+    def test_get_data_vmware_seed_platform_info(self):
+        """Platform info properly reports when on VMware platform."""
+        paths = Paths({'cloud_dir': self.tdir, 'run_dir': self.tdir})
+        # Write ovf-env.xml seed file
+        seed_dir = self.tmp_path('seed', dir=self.tdir)
+        ovf_env = self.tmp_path('ovf-env.xml', dir=seed_dir)
+        util.write_file(ovf_env, OVF_ENV_CONTENT)
+        ds = self.datasource(sys_cfg={}, distro={}, paths=paths)
+
+        self.assertEqual('ovf', ds.cloud_name)
+        self.assertEqual('ovf', ds.platform_type)
+        MPATH = 'cloudinit.sources.DataSourceOVF.'
+        with mock.patch(MPATH + 'util.read_dmi_data', return_value='VMWare'):
+            with mock.patch(MPATH + 'transport_vmware_guestd') as m_guestd:
+                with mock.patch(MPATH + 'transport_iso9660') as m_iso9660:
+                    m_iso9660.return_value = (None, 'ignored', 'ignored')
+                    m_guestd.return_value = (None, 'ignored', 'ignored')
+                    self.assertTrue(ds.get_data())
+                    self.assertEqual(
+                        'vmware (%s/seed/ovf-env.xml)' % self.tdir,
+                        ds.subplatform)
+
 
 class TestTransportIso9660(CiTestCase):
 
diff --git a/tests/unittests/test_datasource/test_smartos.py b/tests/unittests/test_datasource/test_smartos.py
index 46d67b9..42ac697 100644
--- a/tests/unittests/test_datasource/test_smartos.py
+++ b/tests/unittests/test_datasource/test_smartos.py
@@ -426,6 +426,13 @@ class TestSmartOSDataSource(FilesystemMockingTestCase):
         self.assertEqual(MOCK_RETURNS['sdc:uuid'],
                          dsrc.metadata['instance-id'])
 
+    def test_platform_info(self):
+        """All platform-related attributes are properly set."""
+        dsrc = self._get_ds(mockdata=MOCK_RETURNS)
+        self.assertEqual('joyent', dsrc.cloud_name)
+        self.assertEqual('joyent', dsrc.platform_type)
+        self.assertEqual('serial (/dev/ttyS1)', dsrc.subplatform)
+
     def test_root_keys(self):
         dsrc = self._get_ds(mockdata=MOCK_RETURNS)
         ret = dsrc.get_data()