diff options
Diffstat (limited to 'cloudinit/sources/__init__.py')
| -rw-r--r-- | cloudinit/sources/__init__.py | 76 |
1 files changed, 62 insertions, 14 deletions
diff --git a/cloudinit/sources/__init__.py b/cloudinit/sources/__init__.py index a775f1a8..730e8174 100644 --- a/cloudinit/sources/__init__.py +++ b/cloudinit/sources/__init__.py @@ -38,8 +38,12 @@ DEP_FILESYSTEM = "FILESYSTEM" DEP_NETWORK = "NETWORK" DS_PREFIX = 'DataSource' -# File in which instance meta-data, user-data and vendor-data is written +# File in which public available instance meta-data is written +# security-sensitive key values are redacted from this world-readable file INSTANCE_JSON_FILE = 'instance-data.json' +# security-sensitive key values are present in this root-readable file +INSTANCE_JSON_SENSITIVE_FILE = 'instance-data-sensitive.json' +REDACT_SENSITIVE_VALUE = 'redacted for non-root user' # Key which can be provide a cloud's official product name to cloud-init METADATA_CLOUD_NAME_KEY = 'cloud-name' @@ -58,7 +62,7 @@ class InvalidMetaDataException(Exception): pass -def process_instance_metadata(metadata, key_path=''): +def process_instance_metadata(metadata, key_path='', sensitive_keys=()): """Process all instance metadata cleaning it up for persisting as json. Strip ci-b64 prefix and catalog any 'base64_encoded_keys' as a list @@ -67,22 +71,46 @@ def process_instance_metadata(metadata, key_path=''): """ md_copy = copy.deepcopy(metadata) md_copy['base64_encoded_keys'] = [] + md_copy['sensitive_keys'] = [] for key, val in metadata.items(): if key_path: sub_key_path = key_path + '/' + key else: sub_key_path = key + if key in sensitive_keys or sub_key_path in sensitive_keys: + md_copy['sensitive_keys'].append(sub_key_path) if isinstance(val, str) and val.startswith('ci-b64:'): md_copy['base64_encoded_keys'].append(sub_key_path) md_copy[key] = val.replace('ci-b64:', '') if isinstance(val, dict): - return_val = process_instance_metadata(val, sub_key_path) + return_val = process_instance_metadata( + val, sub_key_path, sensitive_keys) md_copy['base64_encoded_keys'].extend( return_val.pop('base64_encoded_keys')) + md_copy['sensitive_keys'].extend( + return_val.pop('sensitive_keys')) md_copy[key] = return_val return md_copy +def redact_sensitive_keys(metadata, redact_value=REDACT_SENSITIVE_VALUE): + """Redact any sensitive keys from to provided metadata dictionary. + + Replace any keys values listed in 'sensitive_keys' with redact_value. + """ + if not metadata.get('sensitive_keys', []): + return metadata + md_copy = copy.deepcopy(metadata) + for key_path in metadata.get('sensitive_keys'): + path_parts = key_path.split('/') + obj = md_copy + for path in path_parts: + if isinstance(obj[path], dict) and path != path_parts[-1]: + obj = obj[path] + obj[path] = redact_value + return md_copy + + URLParams = namedtuple( 'URLParms', ['max_wait_seconds', 'timeout_seconds', 'num_retries']) @@ -127,6 +155,10 @@ class DataSource(object): _dirty_cache = False + # N-tuple of keypaths or keynames redact from instance-data.json for + # non-root users + sensitive_metadata_keys = ('security-credentials',) + def __init__(self, sys_cfg, distro, paths, ud_proc=None): self.sys_cfg = sys_cfg self.distro = distro @@ -152,12 +184,24 @@ class DataSource(object): def _get_standardized_metadata(self): """Return a dictionary of standardized metadata keys.""" - return {'v1': { - 'local-hostname': self.get_hostname(), - 'instance-id': self.get_instance_id(), - 'cloud-name': self.cloud_name, - 'region': self.region, - 'availability-zone': self.availability_zone}} + 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. + return { + 'v1': { + 'availability-zone': availability_zone, + 'availability_zone': availability_zone, + 'cloud-name': cloud_name, + 'cloud_name': cloud_name, + 'instance-id': instance_id, + 'instance_id': instance_id, + 'local-hostname': local_hostname, + 'local_hostname': local_hostname, + 'region': self.region}} def clear_cached_attrs(self, attr_defaults=()): """Reset any cached metadata attributes to datasource defaults. @@ -200,9 +244,7 @@ class DataSource(object): """ instance_data = { 'ds': { - 'meta_data': self.metadata, - 'user_data': self.get_userdata_raw(), - 'vendor_data': self.get_vendordata_raw()}} + 'meta_data': self.metadata}} if hasattr(self, 'network_json'): network_json = getattr(self, 'network_json') if network_json != UNSET: @@ -217,7 +259,9 @@ class DataSource(object): # Process content base64encoding unserializable values content = util.json_dumps(instance_data) # Strip base64: prefix and set base64_encoded_keys list. - processed_data = process_instance_metadata(json.loads(content)) + processed_data = process_instance_metadata( + json.loads(content), + sensitive_keys=self.sensitive_metadata_keys) except TypeError as e: LOG.warning('Error persisting instance-data.json: %s', str(e)) return False @@ -225,7 +269,11 @@ class DataSource(object): LOG.warning('Error persisting instance-data.json: %s', str(e)) return False json_file = os.path.join(self.paths.run_dir, INSTANCE_JSON_FILE) - write_json(json_file, processed_data, mode=0o600) + write_json(json_file, processed_data) # World readable + json_sensitive_file = os.path.join(self.paths.run_dir, + INSTANCE_JSON_SENSITIVE_FILE) + write_json(json_sensitive_file, + redact_sensitive_keys(processed_data), mode=0o600) return True def _get_data(self): |