Skip to content

Microservices can be developed and tested simultaneously. They allow for easier deployment and troubleshooting, making them easier to maintain. Microservices allow small development teams to work almost independently and so they are the perfect choice for scale-up projects.

Notifications You must be signed in to change notification settings

hsnylmz008/golang_microservice_architecture

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

12 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

GoLang Microservice Architecture

Sample architecture with Go, showing how the different pieces are tied together.

Architecture

architecture

TODO: Cleanup the diagram, and add more services

Requirements

Non-functional requirements:

  • system should be highly resilient
  • system should be available
  • system should be scalable
  • system should be observable

Abstract

The following architecture aims to solve several problems

  • Load balancing between services
  • Service discovery and registration
  • Resiliency patterns such as circuit breaker, timeout, retries, rate-limiting
  • Centralized logging
  • Health checks for services
  • Telemetry metrics collection and dashboards
  • blue/green deployment (traffic splitting), rolling upgrades
  • gRPC load balancing and discovery
  • open tracing capabilities
  • demonstrates the delegation of several capabilities to the infra, rather than repeating it at the code levels
  • add block ip functionality
  • add security pipeline
  • add log analysis to process incoming logs
  • add context logging (request id) that propagates through the system
  • enhance/standardize monitoring with opencensus
  • harden security (container user/group)
  • add policy on resources limit (CPU, memory, number of instance etc)

Start

$ docker-compose up -d

UI

Call the Echo Service

Scale the service:

$ docker-compose up -d --scale node=10
$ repeat 10; curl -H "Host: echo.consul.localhost" localhost:80 && printf "\n";

Output:

{"hostname":"e7d4b1cc317c","text":"hello"}
{"hostname":"0bdb052e096a","text":"hello"}
{"hostname":"62fa842dcf9c","text":"hello"}
{"hostname":"2e7f8dcbbc43","text":"hello"}
{"hostname":"1a59218e8121","text":"hello"}
{"hostname":"bee0e7437024","text":"hello"}
{"hostname":"547af30289ee","text":"hello"}
{"hostname":"fec9b78a7e7c","text":"hello"}
{"hostname":"a27f75db0290","text":"hello"}
{"hostname":"5a9726496329","text":"hello"}

Calling Linkerd

In linkerd.yaml, we set it to listen to consul for changes and register the services there. We expose the port :4040 as the load balancer ingress, and set the identifier to io.l5d.header.token. To call the echo service:

$ curl -H "Host: echo" localhost:4140

Setup Namerd

Creating Egress

$ curl -v -XPUT -d @config/namerd.egress.dtab -H "Content-Type: application/dtab" http://localhost:4180/api/1/dtabs/consul_egress

Output:

*   Trying ::1...
* TCP_NODELAY set
* Connected to localhost (::1) port 4180 (#0)
> PUT /api/1/dtabs/consul_egress HTTP/1.1
> Host: localhost:4180
> User-Agent: curl/7.54.0
> Accept: */*
> Content-Type: application/dtab
> Content-Length: 34
>
* upload completely sent off: 34 out of 34 bytes
< HTTP/1.1 204 No Content
<
* Connection #0 to host localhost left intact

Creating Ingress

$ curl -v -XPUT -d @config/namerd.ingress.dtab -H "Content-Type: application/dtab" http://localhost:4180/api/1/dtabs/consul_ingress

Output:

*   Trying ::1...
* TCP_NODELAY set
* Connected to localhost (::1) port 4180 (#0)
> PUT /api/1/dtabs/consul_ingress HTTP/1.1
> Host: localhost:4180
> User-Agent: curl/7.54.0
> Accept: */*
> Content-Type: application/dtab
> Content-Length: 35
>
* upload completely sent off: 35 out of 35 bytes
< HTTP/1.1 204 No Content
<
* Connection #0 to host localhost left intact

Make Call to Linkerd

$ curl -H "Host: echo" localhost:4140

To simulate running traffic

# To simulate running traffic, runs for 120s
$ wrk -c1 -d120 -t1  -H "Host: echo" http://localhost:4140

Check Dtabs

$ curl http://localhost:4180/api/1/dtabs/consul_ingress

Fluentd logging

# Find the httpd endpoint and trigger it
$ repeat 10 curl http://localhost:32857/

TODO

  • cleanup code
  • create kubernetes example

Miscellenaous

# Either
  registrator:
    image: gliderlabs/registrator:vendor
    restart: always
    volumes:
      - /var/run/docker.sock:/tmp/docker.sock
    command: -internal=true -deregister=always -ip=docker.for.mac.localhost -cleanup -tags=registrator consul://consul:8500

# Or
  registrator:
    image: gliderlabs/registrator:vendor
    restart: always
    volumes:
      - /var/run/docker.sock:/tmp/docker.sock
    command: -internal=true -deregister=always -ip=docker.for.mac.localhost -cleanup -tags=registrator consul:8500
    network_mode: host

About

Microservices can be developed and tested simultaneously. They allow for easier deployment and troubleshooting, making them easier to maintain. Microservices allow small development teams to work almost independently and so they are the perfect choice for scale-up projects.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published