DEPRECATED UNTIL SCHEDULER IS UP TO DATE WITH LATEST AWLESS VERSION

The scheduler service is a daemon service that receives templates to be ran and reverted at a later time.

The service basically get templates, validates and stores them. Then it will check every so often when stored templates need to be executed.

go test ./... -v

As an unix sock daemon:

go build; ./awless-scheduler              # default to scheduler service on unix sock

As an HTTP daemon:

go build; ./awless-scheduler --http-mode  # default to scheduler service on localhost:8083
go build; ./awless-scheduler --http-mode --scheduler-hostport 0.0.0.0:9090

Clients use the discovery service to know where the scheduler service is running. By default, the discovery service runs on localhost:8082. To run it on a different port:

./awless-scheduler --discovery-hostport localhost:9090

The scheduler is mostly used together with the awless CLI.

From the CLI you can run one-liner, file or remote template and specify the following flags to run your template at a later date:

• --schedule: indicates the CLI that this template will be send to the service instead of being scheduled.
• --run-in: postpone the execution waiting the run-in duration (using Golang duration notation)
• --revert-in: indicates when to revert this template in case it had a succesfull execution

Examples:

awless create instance name=MyInstance --schedule --run-in 2h --revert-in 4h
awless create instance name=MyInstance --schedule --revert-in 1d

Create a new client giving the discovery URL:

cli, err := client.New("http://127.0.0.1:8082")

Behind the scene, the correct client will be instantiated: a UnixSock client or an HTTP client.

Post a template

err := cli.Post(client.Form{
  Region:   "us-west-1",
  RunIn:    "2m",
  RevertIn: "2h",
  Template: txt,
})

List tasks

tasks, err := cli.List()