cri-meraki-scanning-api-endpoint-server

README

Meraki Scanning API Endpoint

Endpoint to use against Meraki's Scanning API.

It takes the information sent by Cisco's Scanning API and stores it on an S3 bucket. To try to handle the 500ms round trip time required by Cisco, processing the data, and sending it to S3 is done using a queue. This allows to send a response back to Cisco before the data is processed.

A case could be made regarding how the data is read from the request. As of now, we are decoding it as we read it and passing the decoded interface to the queue. Shouldn't it be better just to read the JSON data into a list of bytes ([]byte). I measured both implementations and couldn't see a difference in response times. This could be changed in the future if new evidence arises.

Provides 3 endpoints to fulfill the API requirements:

• GET / : Returns the Meraki validator.
• POST / : Handles the reception of data by the API.
• GET / healthz: Health enpoint to use when presenting the service through an Application Load Balancer.

The service can be run standalone or behind an ALB. Meraki requires that the endpoints are served through SSL, so if you plan to run it in standalone mode, you should configure the ssl, server-key, and server-crt options. You should also configure the port option to 443. Remember to give the binaries the capabilities to listen on ports below 1024. On ubuntu you cand do something like this:

sudo apt-get install libcap2-bin - y
sudo setcap 'cap_net_bind_service=+ep' / your/path/gobinary

[Source](https: // fabianlee.org/2017/05/21/golang-running-a-go-binary-as-a-systemd-service-on-ubuntu-16-04/)

You need valid certificates. If you don't own one, try using [Let's Encrypt](https: // letsencrypt.org).

Installation

Run the binary

Download the binary from the releases site:

[https: // github.com/guzmonne/meraki_endpoint/releases](https: // github.com/guzmonne/meraki_endpoint/releases)

Give the app execution permissions and run it with the appropiate options

cd / tmp
export VERSION = 0.0.3
wget https: // github.com/guzmonne/meraki_endpoint/releases/download /$VERSION/meraki_endpoint
chmod + x meraki_endpoint
./meraki_endpoint \
  - -port= "8080" \
  - -validator= "das..." \
  - -secret= "s3cr3t" \
  - -bucket= "your.s3.bucket.com"

You can see all the app options by running meraki_endpoint - -help.

Run as a service

This steps have only been tested on ** Ubuntu Server 18.04**.

Clone the repo to the server where the service should run. Then, edit the meraki_endpoint.service and configure the options that match your setup(meraki secret, meraki validator, s3 bucket, ssl, etc.) You must configure this options by modifying the value of the ExecStart key.

# ...

WorkingDirectory = /srv/meraki_endpoint
# Configure the necessary options to run the service on your setup
# ExecStart=/usr/bin/meraki_endpoint --port="8080" --bucket="some.bucket" --secret="s3cr3t" --validator="d283..."
# To use it in standalone mode serving http, remember to configure the ssl ans server options besides the other options
# ExecStart=/usr/bin/meraki_endpoint --port="443" --server-crt="/path/to/crt" --server-key="path/to/key"
ExecStart = /usr/bin/meraki_endpoint

# ...

After editing the service file, give execute permissions to the ./install.sh script and run it. You can select a specific version of the service to install by modifying the MERAKI_ENDPOINT_VERSION variable before calling the script.

chmod + x ./install.sh
# To install the stable version
./install.sh
# To install a custom version
MERAKI_ENDPOINT_VERSION = 0.0.3 ./install.sh

If everything goes well, the meraki_endpoint service will be enabled and running. You can check the logs using journalctl.

sudo journalctl - f - u meraki_endpoint

Build

Clone the project into your $GOPATH folder and run go build.

Licence

MIT

Contributions

Welcomed. Create a PR and send it. If it looks good, it will be added to the repository.

Author

Guzmán Monné << godevops@outlook.com>>