# gRPC API
* Table of Contents
{:toc}
[gRPC](https://grpc.io/) is a client/server API for interfacing with CORE
and used by the python GUI for driving all functionality. It is dependent
on having a running `core-daemon` instance to be leveraged.
A python client can be created from the raw generated grpc files included
with CORE or one can leverage a provided gRPC client that helps encapsulate
some of the functionality to try and help make things easier.
## Python Client
A python client wrapper is provided at
[CoreGrpcClient](https://github.com/coreemu/core/blob/master/daemon/core/api/grpc/client.py)
to help provide some conveniences when using the API.
### Client HTTP Proxy
Since gRPC is HTTP2 based, proxy configurations can cause issues. By default
the client disables proxy support to avoid issues when a proxy is present.
You can enable and properly account for this issue when needed.
## Proto Files
Proto files are used to define the API and protobuf messages that are used for
interfaces with this API.
They can be found
[here](https://github.com/coreemu/core/tree/master/daemon/proto/core/api/grpc)
to see the specifics of
what is going on and response message values that would be returned.
## Examples
### Node Models
When creating nodes of type `NodeType.DEFAULT` these are the default models
and the services they map to.
* mdr
* zebra, OSPFv3MDR, IPForward
* PC
* DefaultRoute
* router
* zebra, OSPFv2, OSPFv3, IPForward
* host
* DefaultRoute, SSH
### Interface Helper
There is an interface helper class that can be leveraged for convenience
when creating interface data for nodes. Alternatively one can manually create
a `core.api.grpc.core_pb2.Interface` class instead with appropriate information.
Manually creating gRPC interface data:
```python
from core.api.grpc import core_pb2
# id is optional and will set to the next available id
# name is optional and will default to eth<id>
# mac is optional and will result in a randomly generated mac
iface_data = core_pb2.Interface(
id=0,
name="eth0",
ip4="10.0.0.1",
ip4_mask=24,
ip6="2001::",
ip6_mask=64,
)
```
Leveraging the interface helper class:
```python
from core.api.grpc import client
iface_helper = client.InterfaceHelper(ip4_prefix="10.0.0.0/24", ip6_prefix="2001::/64")
# node_id is used to get an ip4/ip6 address indexed from within the above prefixes
# iface_id is required and used exactly for that
# name is optional and would default to eth<id>
# mac is optional and will result in a randomly generated mac
iface_data = iface_helper.create_iface(
node_id=1, iface_id=0, name="eth0", mac="00:00:00:00:aa:00"
)
```
### Listening to Events
Various events that can occur within a session can be listened to.
Event types:
* session - events for changes in session state and mobility start/stop/pause
* node - events for node movements and icon changes
* link - events for link configuration changes and wireless link add/delete
* config - configuration events when legacy gui joins a session
* exception - alert/error events
* file - file events when the legacy gui joins a session
```python
from core.api.grpc import core_pb2
def event_listener(event):
print(event)
# provide no events to listen to all events
core.events(session_id, event_listener)
# provide events to listen to specific events
core.events(session_id, event_listener, [core_pb2.EventType.NODE])
```
### Configuring Links
Links can be configured at the time of creation or during runtime.
Currently supported configuration options:
* bandwidth (bps)
* delay (us)
* duplicate (%)
* jitter (us)
* loss (%)
```python
from core.api.grpc import core_pb2
# configuring when creating a link
options = core_pb2.LinkOptions(
bandwidth=54_000_000,
delay=5000,
dup=5,
loss=5.5,
jitter=0,
)
core.add_link(session_id, n1_id, n2_id, iface1_data, iface2_data, options)
# configuring during runtime
core.edit_link(session_id, n1_id, n2_id, iface1_id, iface2_id, options)
```
### Peer to Peer Example
```python
# required imports
from core.api.grpc import client
from core.api.grpc.core_pb2 import Node, NodeType, Position, SessionState
# interface helper
iface_helper = client.InterfaceHelper(ip4_prefix="10.0.0.0/24", ip6_prefix="2001::/64")
# create grpc client and connect
core = client.CoreGrpcClient()
core.connect()
# create session and get id
response = core.create_session()
session_id = response.session_id
# change session state to configuration so that nodes get started when added
core.set_session_state(session_id, SessionState.CONFIGURATION)
# create node one
position = Position(x=100, y=100)
n1 = Node(type=NodeType.DEFAULT, position=position, model="PC")
response = core.add_node(session_id, n1)
n1_id = response.node_id
# create node two
position = Position(x=300, y=100)
n2 = Node(type=NodeType.DEFAULT, position=position, model="PC")
response = core.add_node(session_id, n2)
n2_id = response.node_id
# links nodes together
iface1 = iface_helper.create_iface(n1_id, 0)
iface2 = iface_helper.create_iface(n2_id, 0)
core.add_link(session_id, n1_id, n2_id, iface1, iface2)
# change session state
core.set_session_state(session_id, SessionState.INSTANTIATION)
```
### Switch/Hub Example
```python
# required imports
from core.api.grpc import client
from core.api.grpc.core_pb2 import Node, NodeType, Position, SessionState
# interface helper
iface_helper = client.InterfaceHelper(ip4_prefix="10.0.0.0/24", ip6_prefix="2001::/64")
# create grpc client and connect
core = client.CoreGrpcClient()
core.connect()
# create session and get id
response = core.create_session()
session_id = response.session_id
# change session state to configuration so that nodes get started when added
core.set_session_state(session_id, SessionState.CONFIGURATION)
# create switch node
position = Position(x=200, y=200)
switch = Node(type=NodeType.SWITCH, position=position)
response = core.add_node(session_id, switch)
switch_id = response.node_id
# create node one
position = Position(x=100, y=100)
n1 = Node(type=NodeType.DEFAULT, position=position, model="PC")
response = core.add_node(session_id, n1)
n1_id = response.node_id
# create node two
position = Position(x=300, y=100)
n2 = Node(type=NodeType.DEFAULT, position=position, model="PC")
response = core.add_node(session_id, n2)
n2_id = response.node_id
# links nodes to switch
iface1 = iface_helper.create_iface(n1_id, 0)
core.add_link(session_id, n1_id, switch_id, iface1)
iface1 = iface_helper.create_iface(n2_id, 0)
core.add_link(session_id, n2_id, switch_id, iface1)
# change session state
core.set_session_state(session_id, SessionState.INSTANTIATION)
```
### WLAN Example
```python
# required imports
from core.api.grpc import client
from core.api.grpc.core_pb2 import Node, NodeType, Position, SessionState
# interface helper
iface_helper = client.InterfaceHelper(ip4_prefix="10.0.0.0/24", ip6_prefix="2001::/64")
# create grpc client and connect
core = client.CoreGrpcClient()
core.connect()
# create session and get id
response = core.create_session()
session_id = response.session_id
# change session state to configuration so that nodes get started when added
core.set_session_state(session_id, SessionState.CONFIGURATION)
# create wlan node
position = Position(x=200, y=200)
wlan = Node(type=NodeType.WIRELESS_LAN, position=position)
response = core.add_node(session_id, wlan)
wlan_id = response.node_id
# create node one
position = Position(x=100, y=100)
n1 = Node(type=NodeType.DEFAULT, position=position, model="mdr")
response = core.add_node(session_id, n1)
n1_id = response.node_id
# create node two
position = Position(x=300, y=100)
n2 = Node(type=NodeType.DEFAULT, position=position, model="mdr")
response = core.add_node(session_id, n2)
n2_id = response.node_id
# configure wlan using a dict mapping currently
# support values as strings
core.set_wlan_config(session_id, wlan_id, {
"range": "280",
"bandwidth": "55000000",
"delay": "6000",
"jitter": "5",
"error": "5",
})
# links nodes to wlan
iface1 = iface_helper.create_iface(n1_id, 0)
core.add_link(session_id, n1_id, wlan_id, iface1)
iface1 = iface_helper.create_iface(n2_id, 0)
core.add_link(session_id, n2_id, wlan_id, iface1)
# change session state
core.set_session_state(session_id, SessionState.INSTANTIATION)
```
### EMANE Example
For EMANE you can import and use one of the existing models and
use its name for configuration.
Current models:
* core.emane.ieee80211abg.EmaneIeee80211abgModel
* core.emane.rfpipe.EmaneRfPipeModel
* core.emane.tdma.EmaneTdmaModel
* core.emane.bypass.EmaneBypassModel
Their configurations options are driven dynamically from parsed EMANE manifest files
from the installed version of EMANE.
Options and their purpose can be found at the [EMANE Wiki](https://github.com/adjacentlink/emane/wiki).
If configuring EMANE global settings or model mac/phy specific settings, any value not provided
will use the defaults. When no configuration is used, the defaults are used.
```python
# required imports
from core.api.grpc import client
from core.api.grpc.core_pb2 import Node, NodeType, Position, SessionState
from core.emane.ieee80211abg import EmaneIeee80211abgModel
# interface helper
iface_helper = client.InterfaceHelper(ip4_prefix="10.0.0.0/24", ip6_prefix="2001::/64")
# create grpc client and connect
core = client.CoreGrpcClient()
core.connect()
# create session and get id
response = core.create_session()
session_id = response.session_id
# change session state to configuration so that nodes get started when added
core.set_session_state(session_id, SessionState.CONFIGURATION)
# create emane node
position = Position(x=200, y=200)
emane = Node(type=NodeType.EMANE, position=position, emane=EmaneIeee80211abgModel.name)
response = core.add_node(session_id, emane)
emane_id = response.node_id
# create node one
position = Position(x=100, y=100)
n1 = Node(type=NodeType.DEFAULT, position=position, model="mdr")
response = core.add_node(session_id, n1)
n1_id = response.node_id
# create node two
position = Position(x=300, y=100)
n2 = Node(type=NodeType.DEFAULT, position=position, model="mdr")
response = core.add_node(session_id, n2)
n2_id = response.node_id
# configure general emane settings
core.set_emane_config(session_id, {
"eventservicettl": "2"
})
# configure emane model settings
# using a dict mapping currently support values as strings
core.set_emane_model_config(session_id, emane_id, EmaneIeee80211abgModel.name, {
"unicastrate": "3",
})
# links nodes to emane
iface1 = iface_helper.create_iface(n1_id, 0)
core.add_link(session_id, n1_id, emane_id, iface1)
iface1 = iface_helper.create_iface(n2_id, 0)
core.add_link(session_id, n2_id, emane_id, iface1)
# change session state
core.set_session_state(session_id, SessionState.INSTANTIATION)
```
EMANE Model Configuration:
```python
# emane network specific config
core.set_emane_model_config(session_id, emane_id, EmaneIeee80211abgModel.name, {
"unicastrate": "3",
})
# node specific config
core.set_emane_model_config(session_id, node_id, EmaneIeee80211abgModel.name, {
"unicastrate": "3",
})
# node interface specific config
core.set_emane_model_config(session_id, node_id, EmaneIeee80211abgModel.name, {
"unicastrate": "3",
}, iface_id)
```
## Configuring a Service
Services help generate and run bash scripts on nodes for a given purpose.
Configuring the files of a service results in a specific hard coded script being
generated, instead of the default scripts, that may leverage dynamic generation.
The following features can be configured for a service:
* files - files that will be generated
* directories - directories that will be mounted unique to the node
* startup - commands to run start a service
* validate - commands to run to validate a service
* shutdown - commands to run to stop a service
Editing service properties:
```python
# configure a service, for a node, for a given session
core.set_node_service(
session_id,
node_id,
service_name,
files=["file1.sh", "file2.sh"],
directories=["/etc/node"],
startup=["bash file1.sh"],
validate=[],
shutdown=[],
)
```
When editing a service file, it must be the name of `config`
file that the service will generate.
Editing a service file:
```python
# to edit the contents of a generated file you can specify
# the service, the file name, and its contents
core.set_node_service_file(
session_id,
node_id,
service_name,
file_name,
"echo hello",
)
```
## File Examples
File versions of the network examples can be found
[here](https://github.com/coreemu/core/tree/master/daemon/examples/grpc).
These examples will create a session using the gRPC API when the core-daemon is running.
You can then switch to and attach to these sessions using either of the CORE GUIs.