Gracefully handle case where mgmt address is not found on the device, plus other fixes

Author: glennmatthews

netbox_onboarding/onboard.py

@@ -17,7 +17,6 @@
 import re
 import socket
 
-from first import first
 from napalm import get_network_driver
 from napalm.base.exceptions import ConnectionException, CommandErrorException
 
@@ -217,6 +216,9 @@ def get_required_info(self):
         try:
             platform_slug = self.get_platform_slug()
             platform_object = self.get_platform_object_from_netbox(platform_slug=platform_slug)
+            if self.ot.platform != platform_object:
+                self.ot.platform = platform_object
+                self.ot.save()
 
             driver_name = platform_object.napalm_driver
 
@@ -248,17 +250,18 @@ def get_required_info(self):
         # locate the interface assigned with the mgmt_ipaddr value and retain
         # the interface name and IP prefix-length so that we can use it later
         # when creating the IPAM IP-Address instance.
+        # Note that in some cases (e.g., NAT) the mgmt_ipaddr may differ than
+        # the interface addresses present on the device. We need to handle this.
 
-        try:
-            mgmt_ifname, mgmt_pflen = first(
-                (if_name, if_addr_data["prefix_length"])
-                for if_name, if_data in ip_ifs.items()
-                for if_addr, if_addr_data in if_data["ipv4"].items()
-                if if_addr == mgmt_ipaddr
-            )
+        def get_mgmt_info():
+            """Get the interface name and prefix length for the management interface."""
+            for if_name, if_data in ip_ifs.items():
+                for if_addr, if_addr_data in if_data["ipv4"].items():
+                    if if_addr == mgmt_ipaddr:
+                        return (if_name, if_addr_data["prefix_length"])
+            return ("PLACEHOLDER", 0)
 
-        except Exception as exc:
-            raise OnboardException(reason="fail-general", message=str(exc))
+        mgmt_ifname, mgmt_pflen = get_mgmt_info()
 
         # retain the attributes that will be later used by NetBox processing.
 
@@ -322,14 +325,14 @@ def ensure_device_type(
         # instance.
 
         try:
-            self.manufacturer = Manufacturer.objects.get(slug=self.netdev.vendor)
+            self.manufacturer = Manufacturer.objects.get(slug=self.netdev.vendor.lower())
         except Manufacturer.DoesNotExist:
             if not create_manufacturer:
                 raise OnboardException(
                     reason="fail-config", message=f"ERROR manufacturer not found: {self.netdev.vendor}"
                 )
 
-            self.manufacturer = Manufacturer.objects.create(name=self.netdev.vendor, slug=self.netdev.vendor)
+            self.manufacturer = Manufacturer.objects.create(name=self.netdev.vendor, slug=self.netdev.vendor.lower())
             self.manufacturer.save()
 
         # Now see if the device type (slug) already exists,
@@ -343,7 +346,9 @@ def ensure_device_type(
             logging.warning("device model is now: %s", self.netdev.model)
 
         try:
-            self.device_type = DeviceType.objects.get(slug=self.netdev.model)
+            self.device_type = DeviceType.objects.get(slug=self.netdev.model.lower())
+            self.netdev.ot.device_type = self.device_type.slug
+            self.netdev.ot.save()
         except DeviceType.DoesNotExist:
             if not create_device_type:
                 raise OnboardException(
@@ -352,9 +357,11 @@ def ensure_device_type(
 
             logging.info("CREATE: device-type: %s", self.netdev.model)
             self.device_type = DeviceType.objects.create(
-                slug=self.netdev.model, model=self.netdev.model.upper(), manufacturer=self.manufacturer
+                slug=self.netdev.model.lower(), model=self.netdev.model.upper(), manufacturer=self.manufacturer
             )
             self.device_type.save()
+            self.netdev.ot.device_type = self.device_type.slug
+            self.netdev.ot.save()
             return
 
         if self.device_type.manufacturer.id != self.manufacturer.id:
@@ -398,6 +405,7 @@ def ensure_device_role(
 
     def ensure_device_instance(self):
         """Ensure that the device instance exists in NetBox and is assigned the provided device role or DEFAULT_ROLE."""
+        # TODO: this can create duplicate entries in NetBox...
         device, _ = Device.objects.get_or_create(
             name=self.netdev.hostname,
             device_type=self.device_type,
@@ -409,7 +417,7 @@ def ensure_device_instance(self):
         device.serial = self.netdev.serial_number
         device.save()
 
-        self.netdev.ot.device = device
+        self.netdev.ot.created_device = device
         self.netdev.ot.save()
 
         self.device = device

netbox_onboarding/tables.py

@@ -20,10 +20,22 @@ class OnboardingTaskTable(BaseTable):
     """Table for displaying OnboardingTask instances."""
 
     site = tables.LinkColumn()
+    platform = tables.LinkColumn()
+    created_device = tables.LinkColumn()
 
     class Meta(BaseTable.Meta):  # noqa: D106 "Missing docstring in public nested class"
         model = OnboardingTask
-        fields = ("pk", "created_on", "ip_address", "site", "platform", "device", "status", "failed_reason", "message")
+        fields = (
+            "pk",
+            "created_on",
+            "ip_address",
+            "site",
+            "platform",
+            "created_device",
+            "status",
+            "failed_reason",
+            "message",
+        )
 
 
 class OnboardingTaskFeedBulkTable(BaseTable):

netbox_onboarding/tests/test_onboard.py

@@ -117,7 +117,7 @@ def test_ensure_device_instance_not_exist(self):
 
         nbk.ensure_device_instance()
         self.assertIsInstance(nbk.device, Device)
-        self.assertEqual(nbk.device, nbk.netdev.ot.device)
+        self.assertEqual(nbk.device, nbk.netdev.ot.created_device)
         self.assertEqual(nbk.device.serial, "123456")
 
     def test_ensure_interface_not_exist(self):

netbox_onboarding/worker.py

@@ -12,6 +12,7 @@
 limitations under the License.
 """
 import logging
+import time
 
 from django_rq import job
 
@@ -29,7 +30,12 @@ def onboard_device(task_id, credentials):
     username = credentials.username
     password = credentials.password
 
-    ot = OnboardingTask.objects.get(id=task_id)
+    try:
+        ot = OnboardingTask.objects.get(id=task_id)
+    except OnboardingTask.DoesNotExist:
+        # TODO: maybe we started before the DB was done writing it, or maybe it was deleted out from under us?
+        time.sleep(1)
+        ot = OnboardingTask.objects.get(id=task_id)
 
     logging.info("START: onboard device")
 
@@ -54,6 +60,7 @@ def onboard_device(task_id, credentials):
     except Exception as exc:
         ot.status = OnboardingStatusChoices.STATUS_FAILED
         ot.failed_reason = OnboardingFailChoices.FAIL_GENERAL
+        ot.message = str(exc)
         ot.save()
         raise
 

poetry.lock

@@ -10,7 +10,6 @@ packages = [
 [tool.poetry.dependencies]
 python = "^3.6 || ^3.7 || ^3.8"
 invoke = "^1.4.1"
-first = "^2.0.2"
 napalm = "^2.5.0"
 
 [tool.poetry.dev-dependencies]