Lambda function to check specific tcp/ip port availability , and report on it

Optionally, it can record metrics to CloudWatch.

All inputs are either defined as environment variables or as part of event data. Event data will take priority over environment variables

HOSTNAME - url to be checked

PORT - http method to use, defaults to GET

TIMEOUT - timeout when connecting in seconds, defaults to 120s

REPORT_AS_CW_METRICS - set to 1 if you wish to store reported data as CW custom metrics, 0 otherwise, defaults to 1

CW_METRICS_NAMESPACE - if CW custom metrics are being reported, this will determine their namespace, defaults to 'TcpPortCheck'

By default, following properties will be rendered in output Json

Available - 0 or 1

TimeTaken - Time in ms it took to get response from remote server. Default timeout is 2 minutes for http requests.

Lambda function is having no external dependencies by design, so no additional packaging steps are required for deploying it, such as doing pip install [libname]

In order to get some metrics which you can alert on, REPORT_AS_CW_METRICS and CW_METRICS_NAMESPACE environment variables are used. Following metrics will be reported

• Available - 0 or 1, whether response was received in timely manner, indicating problems with network, DNS lookup or server timeout

• TimeTaken - Time taken to fetch response, reported in milliseconds

You can either deploy Lambda manually, or through serverless project. If serverless is being chosen as method of deployments use command below, while making sure that you have setup proper access keys. For more information read here

Serverless framework version used during development is 1.23.0, but it is very likely that later versions will function as well

sls deploy

If you are setting up your Lambda function by hand, make sure it has proper IAM permissions to push Cloud Watch metrics data, and to write to CloudWatch logs

To test function locally with simple Google url (default), run following

sls invoke local  -f portcheck

For more complex example with failed health check, look at test/example.json

$ sls invoke local -f portcheck -p test/example.json 
Socket connect result: 35
Sent data to CloudWatch requestId=:a840f551-afda-11e7-bfe8-532dcf9b380b
Result of checking  example.com:23
{
    "TimeTaken": 5002,
    "Available": "0"
}
{
    "TimeTaken": 5002,
    "Available": "0"
}

Pull requests are welcome to serverless project to deploy CloudWatch rules in order to schedule execution of Http Checking Lambda function.