Troll

Gitlab badges

Github badges

Troll is a very simple webserver returning defined response with configurable delay and a few more features.

INFO

• Report Issues: https://github.com/theztd/troll/issues/new
• Listen port: 8080 (changable via PORT env)
• Health Url: /_healthz/ready.json
• Readines URL: /_healthz/ready.json
• Livenes URL: /_healthz/ready.json
• Metrics URL /_healthz/metrics/

Quick Links

• DEMO application (Try it!!!)

Purpouse

• Testing API endpoint (configurable via YAML).
• Responding with random predefined delay.
• Serve static files from FS.
• Ready to use API backend for FE app prototyping.
• Demo CI/CD with GOlang application

Features:

• Define minimal wait interval for response
• Define document root for serving static content
• Aplication name could be defined
• Listen port could be set via ENV
• Print received json data to log and respond in json
• Log with basic request_id
• 404 page dumping request to log
• Generate 503 randomly (simulate errors)
• Fill RAM with each request (simulate mem leaks)
• Ready delay for testing canary releases and readyness check

Build

env GOOS=target-OS GOARCH=target-architecture go build .

RUN

troll -help

  -fail int
    	Returns 503. Set 1 - 10, where 10 = 100% error rate.
  -fill-ram int
    	Fill ram with each request. Set number in bytes.
  -name string
    	Define custom application name (default "troll")
  -ready-delay int
    	Simulate long application init (seconds).
  -root string
    	Define document root for serving files (default "./public")
  -v2-path string
    	Define path to v2 api endpoint configuration yaml (default "./v2_api.yaml")
  -wait int
    	Minimal wait time before each request

Config

Configuration is possible via ENV variables

Key	Default	description
ADDRESS	:8080	Set 127.0.0.1:8080 to listen only on localhost
LOG_LEVEL	info	Valid options are: error, warning, info, debug

More options are posible via arguments. Configuration of v2 api is possible as follows...

Custom API definition

By editing v2_api.yaml you can change /v2 endpoints and his responses (return code including).

The default structure is:

---
name: Inventory
description: Our company inventory includes employees and equipment
version: 2022-02-09
endpoints:
- path: /employee
  method: GET
  code: 200
  response: "List of our employee..."

Dependencies

Name	Url	Notes
FS:	./v2_api.yaml	v2 config file have to be available.

Monitoring

Health check example

# Content-Type: "application/health+json"
{
     "notes":"Troll is ...",
     "status":"pass",
     "version":"1.0.0"
}

Metrics example

# HELP gin_request_duration the time server took to handle the request.
# TYPE gin_request_duration histogram
gin_request_duration_bucket{uri="",le="0.1"} 300
gin_request_duration_bucket{uri="",le="0.3"} 300
gin_request_duration_bucket{uri="",le="1.2"} 300
gin_request_duration_bucket{uri="",le="5"} 300
gin_request_duration_bucket{uri="",le="10"} 300
gin_request_duration_bucket{uri="",le="+Inf"} 300
gin_request_duration_sum{uri=""} 0.41419708799999966
gin_request_duration_count{uri=""} 300
gin_request_duration_bucket{uri="/v1/:item/*id",le="0.1"} 61
gin_request_duration_bucket{uri="/v1/:item/*id",le="0.3"} 184
gin_request_duration_bucket{uri="/v1/:item/*id",le="1.2"} 300
gin_request_duration_bucket{uri="/v1/:item/*id",le="5"} 300
gin_request_duration_bucket{uri="/v1/:item/*id",le="10"} 300
gin_request_duration_bucket{uri="/v1/:item/*id",le="+Inf"} 300
gin_request_duration_sum{uri="/v1/:item/*id"} 72.53894113400001
gin_request_duration_count{uri="/v1/:item/*id"} 300
gin_request_duration_bucket{uri="/v1/status",le="0.1"} 100
gin_request_duration_bucket{uri="/v1/status",le="0.3"} 100
gin_request_duration_bucket{uri="/v1/status",le="1.2"} 100
gin_request_duration_bucket{uri="/v1/status",le="5"} 100
gin_request_duration_bucket{uri="/v1/status",le="10"} 100
gin_request_duration_bucket{uri="/v1/status",le="+Inf"} 100
gin_request_duration_sum{uri="/v1/status"} 0.003233080999999999
gin_request_duration_count{uri="/v1/status"} 100
# HELP gin_request_total all the server received request num.
# TYPE gin_request_total counter
gin_request_total 700
# HELP gin_request_uv_total all the server received ip num.
# TYPE gin_request_uv_total counter
gin_request_uv_total 1
# HELP gin_response_body_total the server send response body size, unit byte
# TYPE gin_response_body_total counter
gin_response_body_total 322135
# HELP gin_uri_request_total all the server received request num with every uri.
# TYPE gin_uri_request_total counter
gin_uri_request_total{code="200",method="GET",uri="/v1/:item/*id"} 300
gin_uri_request_total{code="200",method="GET",uri="/v1/status"} 100
gin_uri_request_total{code="404",method="GET",uri=""} 300
# HELP go_sync_mutex_wait_total_seconds_total Approximate cumulative time goroutines have spent blocked on a sync.Mutex or sync.RWMutex. This metric is useful for identifying global changes in lock contention. Collect a mutex or block profile using the runtime/pprof package for more detailed contention data.
# TYPE go_sync_mutex_wait_total_seconds_total counter
go_sync_mutex_wait_total_seconds_total 0.000130336
# HELP go_threads Number of OS threads created.
# TYPE go_threads gauge
go_threads 10
# HELP promhttp_metric_handler_requests_in_flight Current number of scrapes being served.
# TYPE promhttp_metric_handler_requests_in_flight gauge
promhttp_metric_handler_requests_in_flight 1
# HELP promhttp_metric_handler_requests_total Total number of scrapes by HTTP status code.
# TYPE promhttp_metric_handler_requests_total counter
promhttp_metric_handler_requests_total{code="200"} 2
promhttp_metric_handler_requests_total{code="500"} 0
promhttp_metric_handler_requests_total{code="503"} 0

Logs

Change log level by env LOG_LEVEL

Format

2023/04/18 16:41:11 [application:nodeName] INFO: Running env devel
2023/04/18 16:41:12 [application:nodeName] DEBUG: Metrics DB migration...
2023/04/18 16:41:12 [application:nodeName] DEBUG: Auth DB migration...
2023/04/18 16:41:12 [application:nodeName] INFO: Waiting for request at address :8080

Backup

• There is nothing to backup