Test dns before ping (#520)

I get negative dns caches a lot here, this will help us ensure dns is there before we start pinging

go.mod

@@ -44,6 +44,7 @@ require (
 	github.com/inancgumus/screen v0.0.0-20190314163918-06e984b86ed3
 	github.com/joho/godotenv v1.3.0
 	github.com/ktrysmt/go-bitbucket v0.9.55
+	github.com/likexian/doh v0.7.1
 	github.com/linode/linodego v1.26.0
 	github.com/microsoftgraph/msgraph-sdk-go v0.61.0
 	github.com/mikesmitty/edkey v0.0.0-20170222072505-3356ea4e686a
@@ -204,6 +205,7 @@ require (
 	github.com/leaanthony/go-ansi-parser v1.0.1 // indirect
 	github.com/leaanthony/gosod v1.0.3 // indirect
 	github.com/leaanthony/slicer v1.5.0 // indirect
+	github.com/likexian/gokit v0.25.15 // indirect
 	github.com/lucasb-eyer/go-colorful v1.2.0 // indirect
 	github.com/magiconair/properties v1.8.7 // indirect
 	github.com/mattn/go-localereader v0.0.1 // indirect

go.sum

@@ -1629,6 +1629,10 @@ github.com/lib/pq v1.10.9 h1:YXG7RB+JIjhP29X+OtkiDnYaXQwpS4JEWq7dtCCRUEw=
 github.com/lib/pq v1.10.9/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
 github.com/liggitt/tabwriter v0.0.0-20181228230101-89fcab3d43de h1:9TO3cAIGXtEhnIaL+V+BEER86oLrvS+kWobKpbJuye0=
 github.com/liggitt/tabwriter v0.0.0-20181228230101-89fcab3d43de/go.mod h1:zAbeS9B/r2mtpb6U+EI2rYA5OAXxsYw6wTamcNW+zcE=
+github.com/likexian/doh v0.7.1 h1:0p75orMxXrfGNAOnOEtl2xZ5T9Ctqi++i/WhN6P6hs8=
+github.com/likexian/doh v0.7.1/go.mod h1:aIEOK197o6nC82iRo5+5L5ws+X7xRCmuer3M4In91WA=
+github.com/likexian/gokit v0.25.15 h1:QjospM1eXhdMMHwZRpMKKAHY/Wig9wgcREmLtf9NslY=
+github.com/likexian/gokit v0.25.15/go.mod h1:S2QisdsxLEHWeD/XI0QMVeggp+jbxYqUxMvSBil7MRg=
 github.com/linode/linodego v1.26.0 h1:2tOZ3Wxn4YvGBRgZi3Vz6dab+L16XUntJ9sJxh3ZBio=
 github.com/linode/linodego v1.26.0/go.mod h1:kD7Bf1piWg/AXb9TA0ThAVwzR+GPf6r2PvbTbVk7PMA=
 github.com/lucasb-eyer/go-colorful v1.2.0 h1:1nnpGOrhyZZuNyfu1QjKiUICQ74+3FNCN69Aj6K7nkY=

pkg/up/deploy.go

@@ -41,7 +41,12 @@ func (ctx *Context) Deploy(commit func() error) error {
 		return err
 	}
 
-	if err := ping(fmt.Sprintf("https://console.%s", ctx.Manifest.Network.Subdomain)); err != nil {
+	subdomain := ctx.Manifest.Network.Subdomain
+	if err := testDns(fmt.Sprintf("console.%s", subdomain)); err != nil {
+		return err
+	}
+
+	if err := ping(fmt.Sprintf("https://console.%s", subdomain)); err != nil {
 		return err
 	}
 

pkg/up/ping.go

@@ -6,23 +6,87 @@ import (
 	"net/http"
 	"time"
 
+	"github.com/likexian/doh"
+	"github.com/likexian/doh/dns"
+
 	"github.com/pluralsh/plural-cli/pkg/utils"
 )
 
+func testDns(domain string) error {
+	ping := fmt.Sprintf("Querying %s...\n", domain)
+	success := "DNS fully resolved, testing if console is functional...\n"
+	return retrier(ping, success, func() error {
+		return doTestDns(domain)
+	})
+}
+
+func doTestDns(domain string) error {
+	ctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
+	defer cancel()
+
+	callDns := func(t dns.Type) error {
+		c := doh.Use(doh.CloudflareProvider, doh.GoogleProvider)
+		rsp, err := c.Query(ctx, dns.Domain(domain), t)
+		if err != nil {
+			return err
+		}
+
+		// close the client
+		c.Close()
+
+		// doh dns answer
+		answer := rsp.Answer
+		if len(answer) > 0 {
+			return nil
+		}
+
+		return fmt.Errorf("dns answer was empty")
+	}
+
+	if err := callDns(dns.TypeA); err == nil {
+		return nil
+	}
+
+	if err := callDns(dns.TypeCNAME); err == nil {
+		return nil
+	}
+
+	if err := callDns(dns.TypeAAAA); err == nil {
+		return nil
+	}
+
+	return fmt.Errorf("could not resolve %s dns domain, you likely need to wait for this to propagate", domain)
+}
+
 func ping(url string) error {
+	ping := fmt.Sprintf("Pinging %s...\n", url)
+	success := "Found status code 200, console up!\n"
+	return retrier(ping, success, func() error {
+		resp, err := http.Get(url)
+		if err == nil && resp.StatusCode == 200 {
+			return nil
+		}
+
+		return fmt.Errorf("Console failed to become ready after 5 minutes, you might want to inspect the resources in the plrl-console namespace")
+	})
+}
+
+func retrier(retryMsg, successMsg string, f func() error) error {
 	done := make(chan bool)
 	ctx, cancel := context.WithTimeout(context.Background(), 5*time.Minute)
+	var resErr error
 	defer cancel()
 
 	go func() {
 		for {
-			fmt.Printf("Pinging %s...\n", url)
-			resp, err := http.Get(url)
-			if err == nil && resp.StatusCode == 200 {
-				utils.Success("Found status code 200, console up!\n")
+			fmt.Printf(retryMsg)
+			err := f()
+			if err == nil {
+				utils.Success(successMsg)
 				done <- true
 				return
 			}
+			resErr = err
 			time.Sleep(10 * time.Second)
 		}
 	}()
@@ -31,6 +95,6 @@ func ping(url string) error {
 	case <-done:
 		return nil
 	case <-ctx.Done():
-		return fmt.Errorf("Console failed to become ready after 5 minutes, you might want to inspect the resources in the plrl-console namespace")
+		return resErr
 	}
 }