In order for the API to be able to run correctly, Prometheus metrics should be enabled in the Oasis node's configuration file config.yml
in the /serverdir/etc/
directory which was set during the installation of the Oasis node. To enable the metrics append this to the end of the file :
metrics:
mode: pull
address: 0.0.0.0:9090
Change :9090
to the port you want prometheus to be exposed at.
Please follow the installation guide to install Node Exporter that will be running along side your node.
Configuring the API involves setting up three config files. This is a strict requirement for the API to be able to run.
Start off by cloning this repository and navigating into it:
git clone https://github.com/SimplyVC/oasis_api_server
cd oasis_api_server
The API can be configured in either of two ways, by using the setup script, or manually.
Please note that you will need to have installed Python before using this script. To install Python, pip and pipenv follow this guide.
You can run the Python script run_setup.py
which will guide you through the configuration process, and ask you to enter the values one by one. This can be done by running the following, starting from the root project directory, oasis_api_server
:
pipenv sync
pipenv run python run_setup.py
Alternatively, for advanced users, you can make a copy of the example_*.ini
files inside the config
folder without the example_
prefix, and manually change the values as required.
This section will guide you through the installation of the API and any of its dependencies.
We recommend that the API is installed on a Linux system, given the simpler installation and running process. Nevertheless, instructions on how to install the API on a Windows system are also provided.
You can either run the API from source or run the API using Docker. This affects what dependencies you will need to install.
Running the API from source only requires you to install Golang.
It is assumed that since this API needs to be run on the same machine as the Oasis node then Golang is already installed, therefore no documentation is provided for it's installation,
After having installed golang you can now run the API as follows from the project directory:
bash run_api.sh
Running the API as a service means that it starts up automatically on boot and restarts automatically if it runs into some issue and stops running. To do so, we recommend the following steps:
# Add a new user to run the API
sudo adduser <USER>
# Grant permissions
sudo chown -R <USER>:<USER> <API_DIR>/ # ownership of api
sudo chmod +x <API_DIR>/run_api.sh # execute permission for runner
The service file will now be created:
# Create the service file
sudo nano /etc/systemd/system/oasis_api_server.service
It should contain the following, replacing <USER>
with the created user's name and the two <API_DIR>
with the API Server's installation directory:
[Unit]
Description=Oasis API Server
After=network.target
StartLimitIntervalSec=0
[Service]
Type=simple
Restart=always
User=<USER>
TimeoutStopSec=90s
WorkingDirectory=<API_DIR>/
ExecStart=/bin/bash <API_DIR>/run_api.sh
[Install]
WantedBy=multi-user.target
Lastly, we will enable and start the alerter service:
sudo systemctl enable oasis_api_server.service
sudo systemctl start oasis_api_server.service
Check out systemctl status oasis_api_server
to confirm that the API is running.
To run the API using Docker, you shouldn't be surprised to find out that you need to install Docker.
You will then obtain the Docker image, make sure that the config files are where they should be, and run everything.
To install Docker on your machine, follow this guide
This part can be done in either of two ways, either by building the Docker image yourself, or by downloading it from Docker Hub.
First start off by cloning this repository:
git clone https://github.com/SimplyVC/oasis_api_server
Then run the following commands to build the image:
cd oasis_api_server
docker build -t simplyvc/oasis_api_server:1.0.7 .
The pre-built Docker image can simply be downloaded by running the following command:
docker pull simplyvc/oasis_api_server:1.0.7
The config files needed by the Docker image are the same as those generated in the Configuring the API
section above.
These config files can be moved to any directory of your choosing <CONFIG_DIR>
.
If you created a new user <USER>
earlier on, set the permissions as follows:
sudo chown -R <USER>:<USER> <CONFIG_DIR>
No further steps are required.
Now that the Docker image is on your machine, and you have written configurations for it, you can run it as follows, where <CONFIG_DIR>
is the full path to the folder containing the previously created config files. Suppose that <CONFIG_DIR> is the full path to the folder containing the previously created config files, <INTERNAL_SOCK_DIR> is the full path on your machine where the internal socket file can be found for the Oasis node (e.g: serverdir/node/), and <PATH_IN_NODE_CONFIG> is the specified location in the API setup where the API will look for the internal socket (e.g: serverdir/node/). The API Server must also be able to find the full path of the Sentry tls_identity_cert.pem
files. Therefore the path <INTERNAL_TLS_DIR> is the full path on your machine where the tls_identity_cert.pem
file can be found, and <PATH_IN_SENTRY_CONFIG> is the specified location in the API setup where the API will look for the tls_identity_cert.pem
file (e.g: serverdir/etc/). Now that the Docker image is on your machine and you have written configuration for it, you can run it as follows:
docker run --network="host" -p 127.0.0.1:8686:8686 \
--mount type=bind,source=<CONFIG_DIR>,target=/app/config/ \
--mount type=bind,source=<INTERNAL_SOCK_DIR>,target=<PATH_IN_NODE_CONFIG> \
--mount type=bind,source=<INTERNAL_TLS_DIR>,target=<PATH_IN_SENTRY_CONFIG> \
-d simplyvc/oasis_api_server:1.0.7
Note: The port after -p
and before the :
is used to route a port from the machine to the internal port of the Docker. If this is changed, any program which refers to the API Docker container must refer to this port.
Example: with 8686
:5367
, the API URL must look like http://1.2.3.4:8686
, but the configured port inside the files must be 5367
, it is suggested to leave them them both as 8686
.
If you wish to make sure that the API is running, the following should return {"result":"pong"}
:
curl -X GET http://localhost:8686/api/ping
If you wish to check the API's connection to a node, you can run the following for some node <NODE>
:
curl -X GET http://localhost:8686/api/pingnode?name=<NODE>
Note
Using curl with multiple queries requires the entire url to be inside double quotation marks as seen in the example below:
curl -X GET "http://localhost:8686/api/registry/nodes?name=<NODE>&height=300"