Initial release of Bulk Firmware Upgrade Manager

Author: TKIPisalegacycipher

examples/bulk_firmware_upgrade_manager.py

@@ -0,0 +1,202 @@
+import meraki
+import datetime
+import time
+
+'''
+Cisco Meraki Bulk Firmware Upgrade Manager
+John M. Kuchta .:|:.:|:.  https://github.com/TKIPisalegacycipher
+This script will pull network IDs from an org and then create asynchronous action batches. Each batch will contain, for 
+each network, an action that will delay the upgrade datetime stamp by X days (configurable). Each batch can contain up 
+to 100 actions, therefore, each batch can modify up to 100 networks. 
+
+As always, you should read the docs before diving in. If you know how these features work, then it will be easier to 
+understand and leverage this tool.
+
+Firmware upgrades endpoint: https://developer.cisco.com/meraki/api-v1/#!get-network-firmware-upgrades
+Action batches: https://developer.cisco.com/meraki/api-v1/#!action-batches-overview
+
+NB: Once you start the script, there are no confirmation prompts or previews, so test in a lab if necessary. 
+
+NB: When the final batch has been submitted, depending on the batch size, it may take a few minutes to finish. Feeling 
+creative? Then try extending this script (using existing code, for the most part) to confirm when the batches are 
+complete. Feeling super creative? Wrap this behind a Flask frontend and have yourself a merry little GUI.
+'''
+
+# init Meraki Python SDK session
+dashboard = meraki.DashboardAPI(suppress_logging=True, single_request_timeout=120)
+
+
+# Configurable options
+# Organization ID. Replace this with your actual organization ID.
+organization_id = 'YOUR ORG ID HERE'  # Use your own organization ID.
+time_delta_in_days = 30  # Max is 1 month per the firmware upgrades endpoint docs
+actions_per_batch = 100  # Max number of actions to submit in a batch. 100 is the maximum. Bigger batches take longer.
+wait_factor = 0.33  # Wait factor for action batches when the action batch queue is full.
+
+# Firmware IDs; not needed for rescheduling, only for upgrading. If you plan to use this for upgrading, then you should
+# first GET the availableVersions IDs and use those here instead, since they have probably changed from the time this
+# was published.
+mx_new_firmware_id = 2128  # Did you update this to your actual FW ID by GETing your availableFirmwareVersions?
+mx_old_firmware_id = 2009  # Did you update this to your actual FW ID by GETing your availableFirmwareVersions?
+
+
+def time_formatter(date_time_stamp):
+	# Basic time formatter to return strings that the API requires
+	formatted_date_time_stamp = date_time_stamp.replace(microsecond=0).isoformat() + 'Z'
+	return formatted_date_time_stamp
+
+
+# Time stamps
+utc_now = datetime.datetime.utcnow()
+utc_future = utc_now + datetime.timedelta(days=time_delta_in_days)
+utc_now_formatted = time_formatter(utc_now)
+utc_future_formatted = time_formatter(utc_future)
+
+
+action_reschedule_existing = {
+	"products": {
+		"appliance":
+			{
+				"nextUpgrade": {
+					"time": utc_future_formatted
+				}
+			}
+	}
+}
+
+# Use this action to schedule a new upgrade. If you do not provide a time param (as shown above), it will execute
+# immediately. IMPORTANT: See API docs for more info before using this.
+action_schedule_new_upgrade = {
+	"products": {
+		"appliance":
+			{
+				"nextUpgrade": {
+					"time": utc_future_formatted,
+					"toVersion": {
+						"id": mx_new_firmware_id
+					}
+				}
+			}
+	}
+}
+
+
+# GET the network list
+networks_list = dashboard.organizations.getOrganizationNetworks(
+	organizationId=organization_id
+)
+
+
+def format_single_action(resource, operation, body):
+	# Combine a single set of batch components into an action
+	action = {
+		"resource": resource,
+		"operation": operation,
+		"body": body
+	}
+
+	return action
+
+
+def create_single_upgrade_action(network_id):
+	# Create a single upgrade action
+	# AB component parts, rename action
+	action_resource = f'/networks/{network_id}/firmwareUpgrades'
+	action_operation = 'update'
+	# Choose whether to reschedule an existing or start a new upgrade
+	action_body = action_schedule_new_upgrade
+
+	upgrade_action = format_single_action(action_resource, action_operation, action_body)
+
+	return upgrade_action
+
+
+def run_an_action_batch(org_id, actions_list, synchronous=False):
+	# Create and run an action batch
+	batch_response = dashboard.organizations.createOrganizationActionBatch(
+		organizationId=org_id,
+		actions=actions_list,
+		confirmed=True,
+		synchronous=synchronous
+	)
+
+	return batch_response
+
+
+def create_action_list(net_list):
+	# Creates a list of actions and returns it
+	# Iterate through the list of network IDs and create an action for each, then collect it
+	list_of_actions = list()
+
+	for network in net_list:
+		# Create the action
+		single_action = create_single_upgrade_action(network['id'])
+		list_of_actions.append(single_action)
+
+	return list_of_actions
+
+
+def batch_actions_splitter(batch_actions):
+	# Split the list of actions into smaller lists of maximum 100 actions each
+	# For each ID in range length of network_ids
+	for i in range(0, len(batch_actions), actions_per_batch):
+		# Create an index range for network_ids of 100 items:
+		yield batch_actions[i:i + actions_per_batch]
+
+
+def action_batch_runner(batch_actions_lists, org_id):
+	# Create an action batch for each list of actions
+	# Store the responses
+	responses = list()
+	number_of_batches = len(batch_actions_lists)
+	number_of_batches_submitted = 0
+	wait_seconds = int(30)
+
+	# Make a batch for each list
+	for batch_action_list in batch_actions_lists:
+		action_batch_queue_checker(org_id)
+		batch_response = run_an_action_batch(org_id, batch_action_list)
+		responses.append(batch_response)
+		number_of_batches_submitted += 1
+
+		# Inform user of progress.
+		print(f'Submitted batch {number_of_batches_submitted} of {number_of_batches}.')
+
+	return responses
+
+
+def action_batch_queue_checker(org_id):
+	all_action_batches = dashboard.organizations.getOrganizationActionBatches(organizationId=org_id)
+	running_action_batches = [batch for batch in all_action_batches if batch['status']['completed'] is False and batch['status']['failed'] is False]
+	total_running_actions = 0
+
+	for batch in running_action_batches:
+		batch_actions = len(batch['actions'])
+		total_running_actions += batch_actions
+
+	wait_seconds = total_running_actions * wait_factor
+
+	while len(running_action_batches) > 4:
+		print(f'There are already five action batches in progress with a total of {total_running_actions} running actions. Waiting {wait_seconds} seconds.')
+		time.sleep(wait_seconds)
+		print('Checking again.')
+
+		all_action_batches = dashboard.organizations.getOrganizationActionBatches(organizationId=org_id)
+		running_action_batches = [batch for batch in all_action_batches if batch['status']['completed'] is False and batch['status']['failed'] is False]
+		total_running_actions = 0
+
+		for batch in running_action_batches:
+			batch_actions = len(batch['actions'])
+			total_running_actions += batch_actions
+
+		wait_seconds = total_running_actions * wait_factor
+
+
+# Create a list of upgrade actions
+upgrade_actions_list = create_action_list(networks_list)
+
+# Split the list into multiple lists of max 100 items each
+upgrade_actions_lists = list(batch_actions_splitter(upgrade_actions_list))
+
+# Run the action batches to clone the networks
+upgraded_networks_responses = action_batch_runner(upgrade_actions_lists, organization_id)