enable httperror to return status code from server

[nitishkumar71]

Description

Return HttpError which provides status code returned by openfaas API. We are only returning HttpError for errors in case of server APIs only.

There are lot of breaking changes in the API.

Motivation and Context

• I have raised an issue to propose this change this is required
  Closes Change from returning error to HttpError #4

How Has This Been Tested?

1. Created a test code script and same has been tested against faas-netes and faasd

package main

import (
	"fmt"
	"log"
	"net/http"
	"net/url"
	"os"
	"time"

	"github.com/openfaas/faas-provider/types"
	"github.com/openfaas/go-sdk"
)

func main() {
	username := os.Getenv("OPENFAAS_USERNAME")
	password := os.Getenv("OPENFAAS_PASSWORD")

	gatewayURL, _ := url.Parse(os.Getenv("OPENFAAS_GATEWAY_URL"))
	auth := &sdk.BasicAuth{
		Username: username,
		Password: password,
	}
	fnName := "test-fn"
	nsName := "openfaas-fn"

	client := sdk.NewClient(gatewayURL, auth, http.DefaultClient)

	// get namespace
	namespaces, err := client.GetNamespaces()
	fmt.Printf("Namespaces: %q\n", namespaces)

	// deploy
	err = client.Deploy(types.FunctionDeployment{
		Service:   fnName,
		Image:     fmt.Sprintf("ttl.sh/nitishkumar71/%s:latest", fnName),
		Namespace: nsName,
	})

	if err != nil {
		log.Fatalf("Deploy Failed: %s", err)
	}

	fmt.Printf("Sleep....\n")
	time.Sleep(15 * time.Second)
	fmt.Printf("Sleep Finish\n")

	fn, err := client.GetFunction(fnName, nsName)
	fmt.Printf("Function: %v\n", fn)
	if err != nil {
		log.Fatalf("GetFunction Failed: %s", err)
	}

	fmt.Println("Scale to Zero")
	err = client.ScaleFunction(fnName, nsName, 0)
	if err != nil {
		log.Fatalf("ScaleFunction Failed: %s", err)
	}

	fmt.Printf("Sleep....\n")
	time.Sleep(15 * time.Second)
	fmt.Printf("Sleep Finish\n")

	// delete
	err = client.DeleteFunction(fnName, nsName)
	if err != nil {
		log.Fatalf("DeleteFunction Failed: %s", err)
	}
	fmt.Print("Function Deleted")
}

Types of changes

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to change)

Checklist:

Commits:

• My commit message has a body and describe how this was tested and why it is required.
• I have signed-off my commits with git commit -s for the Developer Certificate of Origin (DCO)

Code:

• My code follows the code style of this project.
• I have added tests to cover my changes.

Docs:

• My change requires a change to the documentation.
• I have updated the documentation accordingly.

[nitishkumar71]

Branch was re-based against master branch and below script was used to run test against both faas-netes and faasd. Tests were successful against both the faas implementation.

package main

import (
	"context"
	"fmt"
	"log"
	"net/http"
	"net/url"
	"os"
	"time"

	"github.com/openfaas/faas-provider/types"
	"github.com/openfaas/go-sdk"
)

func main() {
	username := os.Getenv("OPENFAAS_USERNAME")
	password := os.Getenv("OPENFAAS_PASSWORD")

	gatewayURL, _ := url.Parse(os.Getenv("OPENFAAS_GATEWAY_URL"))
	auth := &sdk.BasicAuth{
		Username: username,
		Password: password,
	}
	fnName := "test-fn"
	nsName := "openfaas-fn"

	client := sdk.NewClient(gatewayURL, auth, http.DefaultClient)

	// get namespace
	namespaces, err := client.GetNamespaces(context.Background())
	fmt.Printf("Namespaces: %q\n", namespaces)

	// deploy
	err = client.Deploy(context.Background(), types.FunctionDeployment{
		Service:   fnName,
		Image:     fmt.Sprintf("ttl.sh/nitishkumar71/%s:latest", fnName),
		Namespace: nsName,
	})

	if err != nil {
		log.Fatalf("Deploy Failed: %s", err)
	}

	fmt.Printf("Sleep....\n")
	time.Sleep(15 * time.Second)
	fmt.Printf("Sleep Finish\n")

	fn, err := client.GetFunction(context.Background(), fnName, nsName)
	fmt.Printf("Function: %v\n", fn)
	if err != nil {
		log.Fatalf("GetFunction Failed: %s", err)
	}

	fmt.Println("Scale to Zero")
	err = client.ScaleFunction(context.Background(), fnName, nsName, 0)
	if err != nil {
		log.Fatalf("ScaleFunction Failed: %s", err)
	}

	fmt.Printf("Sleep....\n")
	time.Sleep(15 * time.Second)
	fmt.Printf("Sleep Finish\n")

	// delete
	err = client.DeleteFunction(context.Background(), fnName, nsName)
	if err != nil {
		log.Fatalf("DeleteFunction Failed: %s", err)
	}
	fmt.Print("Function Deleted")
}

[alexellis]

I wonder if it's obvious that error is a wrapped HTTPError that can be unwraped?

func (s *Client) deploy(ctx context.Context, method string, spec types.FunctionDeployment) (int, error) {
func (s *Client) deploy(ctx context.Context, method string, spec types.FunctionDeployment) error {

Would it make sense to return HTTPError instead of error given that the int has now been removed?

cc @LucasRoesler

[nitishkumar71]

I wonder if it's obvious that error is a wrapped HTTPError that can be unwraped?

func (s *Client) deploy(ctx context.Context, method string, spec types.FunctionDeployment) (int, error) {
func (s *Client) deploy(ctx context.Context, method string, spec types.FunctionDeployment) error {

Would it make sense to return HTTPError instead of error given that the int has now been removed?

cc @LucasRoesler

We are also returning generic error sometimes. So calling all errors as HttpError, doesn't sound correct to me.

https://github.com/openfaas/go-sdk/pull/8/files#diff-4b667feae66c9d46b21b9ecc19e8958cf4472d162ce0a47ac3e8386af8bbd8cfR219
https://github.com/openfaas/go-sdk/pull/8/files#diff-4b667feae66c9d46b21b9ecc19e8958cf4472d162ce0a47ac3e8386af8bbd8cfR270

[LucasRoesler]

I am not sure i understand the situation here. If the error is caused by some kind of http based error, return an HttpError, if not return the (potentially wrapped) other error. Ideally HttpPError should have an interface that allows you to check for it or you use errors.As to check for it

var httpErr *HttpError
if errors.Is(err, &httpErr) {
   // special http handling
}

See https://pkg.go.dev/errors#example-As

or

type HttpError struct {
	Err    error
	Status int
}

func (e *HttpError) Error() string {
	return e.Err.Error()
}

func (e HttpError) Status() int {
   return e.Status
}

// Then later
type Statuser interface {
    Status() int
}


// then later
if e, ok := err.(Statuser); ok {
    fmt.Printf("we have a status code %d" e.Status())
}

[nitishkumar71]

I am not sure i understand the situation here. If the error is caused by some kind of http based error, return an HttpError, if not return the (potentially wrapped) other error. Ideally HttpPError should have an interface that allows you to check for it or you use errors.As to check for it

var httpErr *HttpError
if errors.Is(err, &httpErr) {
   // special http handling
}

See https://pkg.go.dev/errors#example-As

or

type HttpError struct {
	Err    error
	Status int
}

func (e *HttpError) Error() string {
	return e.Err.Error()
}

func (e HttpError) Status() int {
   return e.Status
}

// Then later
type Statuser interface {
    Status() int
}


// then later
if e, ok := err.(Statuser); ok {
    fmt.Printf("we have a status code %d" e.Status())
}

Agree. This is what I have thought of.