A small go app for health checking endpoints, and collecting other metrics about the base connectivity.
-
Log in to your AWS console
-
Click this link to prepare the stack.
-
Add your specific URL, and your phone number for alerts
-
Click the classic IAM button, and click "Create Stack".
-
Watch all your resources come alive
-
Once the stack is done being created, take down the service! Scale to zero, DDOS yourself, do what you do.
-
Wait for that text message.
Golang has some interesting http tracing data that we can pull from the http library. If a request is successful, we can get more information from the request and post that information to a namespace within AWS CloudWatch. I'd like to deploy this with every one of my applications so that I can create default metrics to health endpoints simply for healthcheck.
As far as the metrics or a custom scraper for those metrics, I'm considering putting that into a different repository.
# reports metrics to cloudwatch
docker run -it --rm -e REGION=us-west-2 -e METRIC_NAME=clustermaestro -e URL="http://my-ip.clustermaestro.com" healthcheck
Created in the AWS AppHealth namespace
| Metric Name | Values | Type | Description |
|---|---|---|---|
is-up |
1 or 0 |
milliseconds (dumb I know) | Is the site up? 0 means no, 1 means 200 response code was returned |
dns-lookup |
0 - n |
milliseconds | How long the dns lookup took to complete |
server-processing |
0 - n |
milliseconds | How long it took for the server to process the request |
tcp-connection |
0 - n |
milliseconds | How long it took for the tcp connection to be established |
tls-handshake |
0 - n |
milliseconds | How long the tls handshake took to accomplish |
total |
0 - n |
milliseconds | Total request time with all that stuff above |