I've been developing with Google App Engine Go for awhile, but I haven't touched it in a year. I tried updating my application from "go1" (1.9?) to "go111", and I'm currently getting some weird errors without any explanation as to what's going on.
The errors include:
The request failed because the instance could not start successfully
Container called exit(1).
500 Internal server error
etc.
None of these point me to any specific line in my code where something would go wrong, nor explain anything more meaningful...
I'm guessing the error is stemming from me upgrading between the golang versions. I had to change the app package into main, add a main function to the application, update the appengine package to a newer version, update the gsuite app, add a cloud compilation widget thingy, change app.yaml script from go to auto, etc.
All in all, I'm lost. A Similar SE question yielded no good answers. Someone else suggested app.yaml might be at fault, so here is mine:
runtime: go111
handlers:
- url: /static
static_dir: static
- url: /res
static_dir: res
- url: /update
script: auto
secure: always
login: admin
- url: /.*
script: auto
secure: always
login: optional
Debug console log is very unhelpful:
And the main file looks essentially like:
package main
import (
"fmt"
"google.golang.org/appengine"
"html/template"
"log"
"net/http"
)
var MainTemplate *template.Template
func main() {
http.HandleFunc("/", hello)
var err error
MainTemplate, err = template.ParseFiles("html/main.html")
if err != nil {
log.Printf("ERROR! %v\n", err.Error())
panic(err)
}
log.Printf("Hello world!\n")
}
func hello(w http.ResponseWriter, r *http.Request) {
c := appengine.NewContext(r)
//................................//
MainTemplate.Execute(w, nil)
}
Anything else it could be?
Okay, after some help from a few comments, here are a few issues that I had to fix to get my code to work:
I converted my init() function into main() without adding a way to listen to requests. The code just ran through the main() function and exited without an error, hence the problems with debugging.
appengine.NewContext(r) is deprecated apparently, so I had to switch those statements to r.Context(). MEANWHILE, Appengine Datastore is still using golang.org/x/net/context and not just context, so if you want to use things like RunInTransaction(), DON'T update your imports, context casts into golang.org/x/net/context just fine
If you follow the official examples provided by Google, you will most likely run into errors like textPayload: "2019/10/20 22:32:46 http: panic serving 127.0.0.1:16051: not an App Engine context. Instead, your code needs to look like the below example.
Since the former app package is now the main package, make sure that any files that are referenced in app.yaml (favicon.ico for example) are in a proper position in relation to the main package (I had to move mine to a different folder to avoid errors popping up every request...).
package main
import (
"google.golang.org/appengine"
"html/template"
"net/http"
)
var MainTemplate *template.Template
func init() {
http.HandleFunc("/", hello)
MainTemplate, _= template.ParseFiles("html/main.html")
}
func main() {
appengine.Main()
}
func hello(w http.ResponseWriter, r *http.Request) {
c := r.Context()
//................................//
MainTemplate.Execute(w, nil)
}
This ought to work. appengine.Main() apparently connects you to all the appengine functionality needed to use the context for datastore / memcache / whatever operations.
Thank you to the commenters that helped me get over the first hump!
Your code is missing the part to start a listener. Add this to your code:
port := os.Getenv("PORT")
if port == "" {
port = "8080"
log.Printf("Defaulting to port %s", port)
}
log.Fatal(http.ListenAndServe(fmt.Sprintf(":%s", port), nil))
Related
There's a problem where I don't know why a context.Context was changed once I pass it to a different package on Google App Engine.
The following code works fine when running on App Engine:
package main
import (
"net/http"
"log"
"google.golang.org/appengine"
)
func main() {
http.HandleFunc("/", myHandler)
appengine.Main()
}
func myHandler(w http.ResponseWriter, r *http.Request) {
ctx := r.Context()
account, err := appengine.ServiceAccount(ctx)
if err != nil {
log.Println("[myHandler] error:", err)
} else {
log.Println("[myHandler] ServiceAccount:", account)
}
w.Write([]byte("ok"))
}
I could retrieve the ServiceAccount successfully when accessing /, and everything was good.
However, when I passed the context from main.go to another package, the function call didn't work. The following was added to main.go:
import (
// other stuff
"github.com/adlerhsieh/q_context/handlers"
)
func main() {
http.HandleFunc("/", myHandler)
appengine.Main()
}
func myHandler(w http.ResponseWriter, r *http.Request) {
ctx := r.Context()
account, err := appengine.ServiceAccount(ctx)
if err != nil {
log.Println("[myHandler] error:", err)
} else {
log.Println("[myHandler] ServiceAccount:", account)
}
handlers.AnotherFunc(ctx) // <--- added this
w.Write([]byte("ok"))
}
Another package:
package handlers
import (
"log"
"context"
"google.golang.org/appengine"
)
func AnotherFunc(ctx context.Context) {
account, err := appengine.ServiceAccount(ctx)
if err != nil {
log.Println("[AnotherFunc] error:", err)
} else {
log.Println("[AnotherFunc] ServiceAccount:", account)
}
}
When I ran it on App Engine, the log said:
2019/09/04 09:36:30 [myHandler] ServiceAccount: myaccount#gmail.com
2019/09/04 09:36:30 [AnotherFunc] error: not an App Engine context
The function calls are the same, but just in different packages. I dug in the package itself and found that it uses the key here (which leads to here) to setup the context. And here to check whether that value was setup properly. However, that value seem to be modified/changed so that the second function call couldn't get it. Even if I omitted the first function call and went straight to the second one, it still has the same error.
Any idea why context object was modified when passing to another package?
The following is my app.yaml:
runtime: go111
service: default
instance_class: F1
automatic_scaling:
min_idle_instances: 0
max_idle_instances: automatic
min_pending_latency: automatic
max_pending_latency: automatic
max_concurrent_requests: 30
handlers:
- url: /.*
script: auto
login: admin
nobuild_files:
- vendor
env_variables:
ENV: 'dev'
GO111MODULE: 'off'
Here is the GitHub repo link.
Thank you!
It turns out that my code actually worked. It's because of some other operation error.
However, I'll just post the issue that actually caused it so it can help those who have the same issue.
With the new go111 runtime, it treats packages from non-root directory or its subdirectories as a different type of package. This caused the problem with "not an App Engine context". I'll just call it an "outcast" package for now (cause I'm not entirely sure why's that).
For example:
- appengine
- main.go
- handlers
- handlers.go <-- this is a regular package
- appengine
- main.go
- handlers
- handlers.go < -- this is an outcast package
An outcast package would have issues handling context.Context generated from App Engine, as pointed out in my question.
The mechanism of App Engine knowing that the context is created from App Engine, is using a built-in value that can only be retrieved from its internal package (with an un-exported pointer-string key). When passing the context to an outcast package, we can no longer retrieve the value from the context. It's still a mystery for me that why the value disappeared, but it's probably because of some Go compiling mechanism.
The solution would be moving the main.go to the top-level directory in the project, so that there would be no outcast package anywhere.
I'm using Google App Engine Standard Environment with the Go 1.11 runtime. The documentation for Go 1.11 says "Write your application logs using stdout for output and stderr for errors". The migration from Go 1.9 guide also suggests not calling the Google Cloud Logging library directly but instead logging via stdout.
https://cloud.google.com/appengine/docs/standard/go111/writing-application-logs
With this in mind, I've written a small HTTP Service (code below) to experiment logging to Stackdriver using JSON output to stdout.
When I print plain text messages they appear as expected in the Logs Viewer panel under textPayload. When I pass a JSON string they appear under jsonPayload. So far, so good.
So, I added a severity field to the output string and Stackdriver Log Viewer successfully categorizes the message according to the levelled logging NOTICE, WARNING etc.
https://cloud.google.com/logging/docs/reference/v2/rest/v2/LogEntry
The docs say to set the trace identifier to correlate log entries with the originating request log. The trace ID is extracted from the X-Cloud-Trace-Context header set by the container.
Simulate it locally using curl -v -H 'X-Cloud-Trace-Context: 1ad1e4f50427b51eadc9b36064d40cc2/8196282844182683029;o=1' http://localhost:8080/
However, this does not cause the messages to be threaded by request, but instead the trace property appears in the jsonPayload object in the logs. (See below).
Notice that severity has been interpreted as expected and does not appear in the jsonPayload. I had expected the same to happen for trace, but instead it appears to be unprocessed.
How can I achieve nested messages within the original request log message? (This must be done using stdout on Go 1.11 as I do not wish to log directly with the Google Cloud logging package).
What exactly is GAE doing to parse the stdout stream from my running process? (In the setup guide for VMs on GCE there is something about installing an agent program to act as a conduit to Stackdriver logging- is this what GAE has installed?)
app.yaml file looks like this:
runtime: go111
handlers:
- url: /.*
script: auto
package main
import (
"fmt"
"log"
"net/http"
"os"
"strings"
)
var projectID = "glowing-market-234808"
func parseXCloudTraceContext(t string) (traceID, spanID, traceTrue string) {
slash := strings.Index(t, "/")
semi := strings.Index(t, ";")
equal := strings.Index(t, "=")
return t[0:slash], t[slash+1 : semi], t[equal+1:]
}
func sayHello(w http.ResponseWriter, r *http.Request) {
xTrace := r.Header.Get("X-Cloud-Trace-Context")
traceID, spanID, _ := parseXCloudTraceContext(xTrace)
trace := fmt.Sprintf("projects/%s/traces/%s", projectID, traceID)
warning := fmt.Sprintf(`{"trace":"%s","spanId":"%s", "severity":"WARNING","name":"Andy","age":45}`, trace, spanID)
fmt.Fprintf(os.Stdout, "%s\n", warning)
message := "Hello"
w.Write([]byte(message))
}
func main() {
http.HandleFunc("/", sayHello)
port := os.Getenv("PORT")
if port == "" {
port = "8080"
}
log.Fatal(http.ListenAndServe(fmt.Sprintf(":%s", port), nil))
}
Output shown in the Log viewer:
...,
jsonPayload: {
age: 45
name: "Andy"
spanId: "13076979521824861986"
trace: "projects/glowing-market-234808/traces/e880a38fb5f726216f94548a76a6e474"
},
severity: "WARNING",
...
I have solved this by adjusting the program to use logging.googleapis.com/trace in place of trace and logging.googleapis.com/spanId in place of spanId.
warning := fmt.Sprintf(`{"logging.googleapis.com/trace":"%s","logging.googleapis.com/spanId":"%s", "severity":"WARNING","name":"Andy","age":45}`, trace, spanID)
fmt.Fprintf(os.Stdout, "%s\n", warning)
It seems that GAE is using the logging agent google-fluentd (a modified version of the fluentd log data colletor.)
See this link for a full explanation.
https://cloud.google.com/logging/docs/agent/configuration#special-fields
[update] June 25th, 2019: I've written a Logrus plugin that will help to thread log entries by HTTP request. It's available under on GitHub https://github.com/andyfusniak/stackdriver-gae-logrus-plugin.
[update]] April 3rd, 2020: I've since switched to using Cloud Run and the Logrus plugin appears to work fine with this platform also.
I've been utilizing the url provided to me by the google-app-engine, i.e. "projectname-id.appspot.com" for my project. I had previously purchased a custom domain from GoDaddy, and I followed the steps at the following link to verify ownership of the domain name within the Google Cloud Platform, add the domain to my Google App Engine Project, and update my DNS settings in GoDaddy to point to the listed CNAME/server combination. https://cloud.google.com/appengine/docs/go/console/using-custom-domains-and-ssl
When I visit my custom domain, a website page is served, but it's always a 404. The "projectname-id.appspot.com" url still works correctly, and when I look at the log statements in the Google App Engine, it receives the request from both my custom domain and the appspot domain - which seems to suggest the domain dns was updated properly. See image below, where the 404 is from the custom domain and the 200 is from the appspot url:
Google App Engine Logs
Is there anything else I have to do? The backend is written in Go, and we're using the Mux Router. Do I need to modifiy my app.yaml file or edit my routes somehow? Any suggestions would be greatly appreciated.
Here I've included some code snippets I use to initialize my server:
App.yaml
version: alpha-001
runtime: go
api_version: go1
handlers:
- url: /.*
script: _go_app
env_variables:
PRODUCTION: 'TRUE'
DATASTORE_DATASET: 'app-id'
DATASTORE_HOST: 'http://localhost:8043'
DATASTORE_EMULATOR_HOST: 'localhost:8043'
DATASTORE_PROJECT_ID: 'app-id'
GOOGLE_APPLICATION_CREDENTIALS: './app-string.json'
Initializing the WebConsole (Server):
func init() {
// Web Server for API Endpoints
flag.Parse()
var server *web_console.WebConsole
if prod := os.Getenv("PRODUCTION"); prod == "FALSE" {
server = web_console.NewWebConsole(false)
} else {
server = web_console.NewWebConsole(true)
}
server.Run()
}
WebConsole.go
type WebConsole struct {
prod bool
Mux *mux.Router
DbMap *gorp.DbMap
Client *datastore.Client
}
func NewWebConsole(prod bool) *WebConsole {
return &WebConsole{
prod: prod,
}
}
func (w *WebConsole) Run() {
w.dbInit()
w.routesInit()
}
func (w *WebConsole) dbInit() {
// Configure SQL connection
// Code removed for privacy reasons
}
func (w *WebConsole) routesInit() {
// Configure routes with Mux
w.Mux = mux.NewRouter()
api.AddCertChallengeApis(w.Mux)
// The path "/" matches everything not matched by some other path.
// Checkout: http://stackoverflow.com/questions/26581231/google-cloud-go-handler-without-router-gorilla-mux
// for more details
http.Handle("/", w.Mux)
}
Api Package File for Routing
package api
import (
"github.com/gorilla/mux"
"google.golang.org/appengine"
"google.golang.org/appengine/log"
"net/http"
"strings"
)
func AddCertChallengeApis(r *mux.Router) {
r.Schemes("http")
r.HandleFunc("/", defaultHandler())
}
func defaultHandler() http.HandlerFunc {
return func(rw http.ResponseWriter, req *http.Request) {
// Construct new app engine context
c := appengine.NewContext(req)
log.Infof(c, "App ID: %v", appengine.AppID(c))
rw.Header().Set("Content-Type", "text/plain")
rw.Write([]byte("hi, welcome to my website yo"))
log.Infof(c, "Hit the website")
}
}
I have done custom domains on appengine/go before, and there is nothing else you should have to do. I would try removing this line, though,
r.Schemes("http")
in case it is something https related.
package helloworld
import (
"fmt"
"net/http"
"appengine"
"appengine/user"
)
func init() {
fmt.Print("hello")
http.HandleFunc("/", handler)
}
func handler(w http.ResponseWriter, r *http.Request) {
c := appengine.NewContext(r)
u := user.Current(c)
if u == nil {
url, err := user.LoginURL(c, r.URL.String())
if err != nil {
http.Error(w, err.Error(), http.StatusInternalServerError)
return
}
w.Header().Set("Location", url)
w.WriteHeader(http.StatusFound)
return
}
fmt.Fprintf(w, "Hello, %v!", u)
}
Throw the following error in goapp serve output
(saucy)adam#localhost:~/projects/ringed-land-605/default$ goapp serve -host 0.0.0.0 .
INFO 2014-06-08 23:57:47,862 devappserver2.py:716] Skipping SDK update check.
INFO 2014-06-08 23:57:47,877 api_server.py:171] Starting API server at: http://localhost:48026
INFO 2014-06-08 23:57:47,923 dispatcher.py:182] Starting module "default" running at: http://0.0.0.0:8080
INFO 2014-06-08 23:57:47,925 admin_server.py:117] Starting admin server at: http://localhost:8000
ERROR 2014-06-08 23:57:48,759 http_runtime.py:262] bad runtime process port ['hello46591\n']
Removing the fmt.Print() fixes the issue. My question is why does that happen?
When starting the runtime process, the Go Development Server (in the App Engine Go SDK) is reading the single line response found in your helloworld's init.
This modifies the _start_process_flavor flag in http_runtime.py; consequently, the HTTP runtime attempts to read the line for direction on which port to listen to.
Read the single line response expected in the start process file. [...] The START_PROCESS_FILE flavor uses a file for the runtime instance to report back the port it is listening on.
In this case, hello isn't a valid port to listen on.
Try using Go's log package instead:
log.Print("hello")
I am trying to test an AppEngine/Go application. I start dev_appserver.py and it begins serving the application, but when I go to localhost:8080 in my browser, I get:
Compile error:
/home/adam/foobar/server/app/server.go:5: can't find import: appengine/users
2011/08/23 19:45:34 go-app-builder: Failed building app: failed running 8g: exit status 1
I feel as though I need to do something to make the AppEngine-specific libraries available where GO expects them to be, but I don't really want to run goinstall on everything that comes in the AppEngine/Go SDK zip, do I? I seem to have missed an installation step, but for the life of me, I can't figure the sane and right thing to do.
I am on Ubuntu, if that matters.
The Users API isn't appengine/users - it's appengine/user. From the example on the App Engine page:
import (
"appengine"
"appengine/user"
)
func welcome(w http.ResponseWriter, r *http.Request) {
c := appengine.NewContext(r)
u := user.Current(c)
if u == nil {
url := u.LoginURL(c, "/")
fmt.Fprintf(w, `Sign in or register`, url)
return
}
url := user.LogoutURL(c, "/")
fmt.Fprintf(w, `Welcome, %s! (sign out)`, u, url)
}
You don't have to compile the code yourself - just run the dev_appserver and it will compile it for you whenever the code changes. Have you gone through the getting started docs?