sources/azure: report success to host and introduce kvp module (#2141)

Add success reporting to the host via KVP.

- Move _report_failure_to_host() into kvp module.
- Tweak error description to use result=error instead of
  PROVISIONING_ERROR: ...
- Use result=success for the successful ("ready") reports.
- report_x_via_kvp => report_x_to_host for consistency with fabric.
  ReportableError.as_description() => as_encoded_report()

Signed-off-by: Chris Patterson <cpatterson@microsoft.com>

4 files changed, 86 insertions, 34 deletions

diff --git a/cloudinit/sources/DataSourceAzure.py b/cloudinit/sources/DataSourceAzure.py
index 969cb376..923644e1 100644
--- a/cloudinit/sources/DataSourceAzure.py
+++ b/cloudinit/sources/DataSourceAzure.py
@@ -25,8 +25,8 @@ from cloudinit.net.dhcp import (
     NoDHCPLeaseMissingDhclientError,
 )
 from cloudinit.net.ephemeral import EphemeralDHCPv4
-from cloudinit.reporting import events, handlers, instantiated_handler_registry
-from cloudinit.sources.azure import errors, identity, imds
+from cloudinit.reporting import events
+from cloudinit.sources.azure import errors, identity, imds, kvp
 from cloudinit.sources.helpers import netlink
 from cloudinit.sources.helpers.azure import (
     DEFAULT_WIRESERVER_ENDPOINT,
@@ -1175,20 +1175,6 @@ class DataSourceAzure(sources.DataSource):
         return reprovision_data
 
     @azure_ds_telemetry_reporter
-    def _report_failure_to_host(self, error: errors.ReportableError) -> bool:
-        """Report failure to host via well-known key."""
-        value = error.as_description()
-        kvp_handler = instantiated_handler_registry.registered_items.get(
-            "telemetry"
-        )
-        if not isinstance(kvp_handler, handlers.HyperVKvpReportingHandler):
-            LOG.debug("KVP handler not enabled, skipping host report.")
-            return False
-
-        kvp_handler.write_key("PROVISIONING_REPORT", value)
-        return True
-
-    @azure_ds_telemetry_reporter
     def _report_failure(self, error: errors.ReportableError) -> bool:
         """Tells the Azure fabric that provisioning has failed.
 
@@ -1196,9 +1182,10 @@ class DataSourceAzure(sources.DataSource):
         @return: The success status of sending the failure signal.
         """
         report_diagnostic_event(
-            f"Azure datasource failure occurred: {error.as_description()}",
+            f"Azure datasource failure occurred: {error.as_encoded_report()}",
             logger_func=LOG.error,
         )
+        kvp.report_failure_to_host(error)
 
         if self._is_ephemeral_networking_up():
             try:
@@ -1253,6 +1240,8 @@ class DataSourceAzure(sources.DataSource):
 
         :returns: List of SSH keys, if requested.
         """
+        kvp.report_success_to_host()
+
         try:
             data = get_metadata_from_fabric(
                 endpoint=self._wireserver_endpoint,
diff --git a/cloudinit/sources/azure/errors.py b/cloudinit/sources/azure/errors.py
index 1a452635..5c4ad7db 100644
--- a/cloudinit/sources/azure/errors.py
+++ b/cloudinit/sources/azure/errors.py
@@ -8,7 +8,7 @@ import logging
 import traceback
 from datetime import datetime
 from io import StringIO
-from typing import Any, Dict, Optional
+from typing import Any, Dict, List, Optional
 
 from cloudinit import version
 from cloudinit.sources.azure import identity
@@ -16,6 +16,22 @@ from cloudinit.sources.azure import identity
 LOG = logging.getLogger(__name__)
 
 
+def encode_report(
+    data: List[str], delimiter: str = "|", quotechar: str = "'"
+) -> str:
+    """Encode report data with csv."""
+    with StringIO() as io:
+        csv.writer(
+            io,
+            delimiter=delimiter,
+            quotechar=quotechar,
+            quoting=csv.QUOTE_MINIMAL,
+        ).writerow(data)
+
+        # strip trailing \r\n
+        return io.getvalue().rstrip()
+
+
 class ReportableError(Exception):
     def __init__(
         self,
@@ -39,10 +55,11 @@ class ReportableError(Exception):
         except Exception as id_error:
             self.vm_id = f"failed to read vm id: {id_error!r}"
 
-    def as_description(
-        self, *, delimiter: str = "|", quotechar: str = "'"
+    def as_encoded_report(
+        self,
     ) -> str:
         data = [
+            "result=error",
             f"reason={self.reason}",
             f"agent={self.agent}",
         ]
@@ -53,18 +70,7 @@ class ReportableError(Exception):
             f"documentation_url={self.documentation_url}",
         ]
 
-        with StringIO() as io:
-            csv.writer(
-                io,
-                delimiter=delimiter,
-                quotechar=quotechar,
-                quoting=csv.QUOTE_MINIMAL,
-            ).writerow(data)
-
-            # strip trailing \r\n
-            csv_data = io.getvalue().rstrip()
-
-        return f"PROVISIONING_ERROR: {csv_data}"
+        return encode_report(data)
 
     def __eq__(self, other) -> bool:
         return (
@@ -75,7 +81,7 @@ class ReportableError(Exception):
         )
 
     def __repr__(self) -> str:
-        return self.as_description()
+        return self.as_encoded_report()
 
 
 class ReportableErrorUnhandledException(ReportableError):
diff --git a/cloudinit/sources/azure/kvp.py b/cloudinit/sources/azure/kvp.py
new file mode 100644
index 00000000..735c4616
--- /dev/null
+++ b/cloudinit/sources/azure/kvp.py
@@ -0,0 +1,57 @@
+# Copyright (C) 2022 Microsoft Corporation.
+#
+# This file is part of cloud-init. See LICENSE file for license information.
+
+import logging
+from datetime import datetime
+from typing import Optional
+
+from cloudinit import version
+from cloudinit.reporting import handlers, instantiated_handler_registry
+from cloudinit.sources.azure import errors, identity
+
+LOG = logging.getLogger(__name__)
+
+
+def get_kvp_handler() -> Optional[handlers.HyperVKvpReportingHandler]:
+    """Get instantiated KVP telemetry handler."""
+    kvp_handler = instantiated_handler_registry.registered_items.get(
+        "telemetry"
+    )
+    if not isinstance(kvp_handler, handlers.HyperVKvpReportingHandler):
+        return None
+
+    return kvp_handler
+
+
+def report_via_kvp(report: str) -> bool:
+    """Report to host via PROVISIONING_REPORT KVP key."""
+    kvp_handler = get_kvp_handler()
+    if kvp_handler is None:
+        LOG.debug("KVP handler not enabled, skipping host report.")
+        return False
+
+    kvp_handler.write_key("PROVISIONING_REPORT", report)
+    return True
+
+
+def report_failure_to_host(error: errors.ReportableError) -> bool:
+    return report_via_kvp(error.as_encoded_report())
+
+
+def report_success_to_host() -> bool:
+    try:
+        vm_id = identity.query_vm_id()
+    except Exception as id_error:
+        vm_id = f"failed to read vm id: {id_error!r}"
+
+    report = errors.encode_report(
+        [
+            "result=success",
+            f"agent=Cloud-Init/{version.version_string()}",
+            f"timestamp={datetime.utcnow().isoformat()}",
+            f"vm_id={vm_id}",
+        ]
+    )
+
+    return report_via_kvp(report)
diff --git a/cloudinit/sources/helpers/azure.py b/cloudinit/sources/helpers/azure.py
index 2413d6b0..6e5c1f43 100644
--- a/cloudinit/sources/helpers/azure.py
+++ b/cloudinit/sources/helpers/azure.py
@@ -1023,7 +1023,7 @@ def get_metadata_from_fabric(
 @azure_ds_telemetry_reporter
 def report_failure_to_fabric(endpoint: str, error: "errors.ReportableError"):
     shim = WALinuxAgentShim(endpoint=endpoint)
-    description = error.as_description()
+    description = error.as_encoded_report()
     try:
         shim.register_with_azure_and_report_failure(description=description)
     finally: