A python module used to interact with the Kea DHCP API daemons (dhcp4, dhcp6, ctrl-agent and ddns) with Pydantic support so your editor support should be pretty, and also provides basic data validation for any models implemented (eg. Subnet4).
pip install pykeadhcp
from pykeadhcp import Kea
server = Kea(host="http://localhost", port=8000)
subnets_v4 = server.dhcp4.subnet4_list()
for subnet in subnets_v4:
print(subnet.subnet, subnet.option_data, subnet.relay, subnet.pools_list)
my_subnet = server.dhcp6.subnet4_get(name="pykeadhcp-pytest")
print(my_subnet.json(exclude_none=True, indent=4))
# {
# "valid_lifetime": 4000,
# "renew_timer": 1000,
# "rebind_timer": 2000,
# "option_data": [],
# "calculate_tee_times": true,
# "t1_percent": 0.5,
# "t2_percent": 0.8,
# "store_extended_info": false,
# "name": "pykeadhcp-pytest",
# "relay": {
# "ip-addresses": []
# },
# "subnet6": [
# {
# "valid_lifetime": 4000,
# "renew_timer": 1000,
# "rebind_timer": 2000,
# "option_data": [],
# "calculate_tee_times": true,
# "t1_percent": 0.5,
# "t2_percent": 0.8,
# "store_extended_info": false,
# "id": 40123,
# "subnet": "2001:db8::/64",
# "preferred_lifetime": 3600,
# "pd_pools": [],
# "rapid_commit": false
# }
# ],
# "rapid_commit": false
# }
from pykeadhcp.models.dhcp4.subnet import Subnet4
my_subnet = Subnet4(
id=1234, subnet="192.0.2.32/31", option_data=[{"code": 3, "data": "192.0.2.32"}]
)
create_subnet = server.dhcp4.subnet4_add(subnets=[my_subnet])
print(create_subnet.result, create_subnet.text)
# Note because subnet_cmds hook library is not loaded, we run into an exception here:
# pykeadhcp.exceptions.KeaHookLibraryNotConfiguredException: Hook library 'subnet_cmds' is not configured for 'dhcp4' service. Please ensure this is enabled in the configuration for the 'dhcp4' daemon
The super handy feature with using Pydantic for all the data models is the fact it uses Python type annotation which most editors/IDEs should be able to provide autocomplete functionality like this:
If you have basic authentication enabled on your Kea Servers, initialize the Kea
class like this:
from pykeadhcp import Kea
server = Kea(host="http://localhost", port=8000, use_basic_auth=True, username="your-username", password="your-password")
Once you initialize the Kea class, it will automatically attempt to gather the configuration for all daemons and cache them locally as cached_config
eg. like:
from pykeadhcp import Kea
server = Kea(host="http://localhost", port=8000)
# Control Agent
print(server.ctrlagent.cached_config)
# Dhcp4
print(server.dhcp4.cached_config)
# Dhcp6
print(server.dhcp6.cached_config)
# DDNS
print(server.ddns.cached_config)
If you make a change via the API that amends the configuration (eg. network4-add), the cached config must be refreshed manually using:
server.dhcp4.refresh_cached_config()
For API calls that don't amend the configuration (eg. lease4-add, lease6-add, config-get, etc....), there is no need to refresh the relevant daemon configuration. Maybe I will add a feature in the future to allow the user to specify if they want the cached_config to be automatically refreshed when a function is called that requires a refresh but for now its manual.
Majority if not all remote-
commands which interact with the configuration backend implement the remote_map
variable which allows you to specify the database instance you want to interact with as per documentation here. It accepts the following variables inside the dictionary:
{
"type": "<mysql | postgresql>",
"host": "<ip address>",
"port": <port>
}
For example:
kea_server.dhcp4.remote_subnet4_set(subnet=subnet, server_tags=["all"], remote_map={"type": "mysql"})
Note that the remote_map is optional and not mandatory.
Majority of the useful API functionality requires a subscription for the premium hooks to use API commands like subnet4-add
or network6-add
as an example. This is typically the recommended way to interact with Kea as there is some additional validation that the API will perform and potentially prevent misconfiguration of your daemons (Dhcp4, Dhcp6, Control Agent, DDNS) that pykeadhcp may not correctly implement.
However with the use of config-test
and config-set
API commands which are supported for all daemons in the free version of ISC Kea DHCP, you can utilize the various parsers implemented in pykeadhcp which provides similar functionality to create, read, update and delete various resources like subnets, shared networks, reservations etc... The parsers attempt to parse a local configuration (eg. server.dhcp4.cached_config) and should be used with the 2 API commands config-test
and config-set
. The problem with config-set
is that it will restart the daemon on the server unlike the specific commands to directly interact with the configuration.
Before continuing, these parsers are manually crafted and can break at anytime when ISC release an update that changes the behaviour of the configuration model. Therefore you should use parsers at your own risk.
from pykeadhcp import Kea
from pykeadhcp.parsers.dhcp4 import Dhcp4Parser
cidr_to_find = "192.168.1.0/24"
server = Kea(host="http://localhost", port=8000)
parser = Dhcp4Parser(config=server.dhcp4.cached_config)
subnet = parser.get_subnet_by_cidr(cidr=cidr_to_find)
if not subnet:
exit(f"Unable to find subnet: {cidr_to_find}")
print(f"Found Subnet {cidr_to_find} (Subnet ID: {subnet.id})")
option_data = subnet.option_data
for option in option_data:
print(f"Code: {option.code}")
print(f"Data: {option.data}")
# Output
# Found Subnet 192.168.1.0/24 (Subnet ID: 1)
# Code: 3
# Data: 192.168.1.1
# Code: 6
# Data: 1.1.1.1,9.9.9.9
All supported commands by the daemons are in the format of the API referenced commands with the exception of replacing any hyphen or space with an underscore. Eg. the build-report
API command for all daemons is implemented as build_report
so it heavily ties into the Kea predefined commands when looking at their documentation. Currently everything is built towards Kea 2.2.0. Pydantic variables will replace any hyphens with an underscore however when loading/exporting the data models, it will replace all keys with the hyphen to adhere to the Kea expected variables, ensure that the KeaBaseModel
(located in from pykeadhcp.models.generic.base import KeaBaseModel
instead of from pydantic import BaseModel
) is used when creating any Pydantic models to inherit this functionality.
See this document which explains the development/setup to contribute to this project
See this document which shows what commands are implemented and tested in the latest release