Env:nodejs12
Folder structure:
#root
/functions
/src
...
/models
/resolvers
index.ts
ormconfig.json
package.json
tsconfig.json
...
.firebaserc
firebase.json
Everything worked when developing in local environment. After deployed to firebase functions, No connection options were found in any orm configuration files shows up.
What might be the cause?
I'll update with more information if needed.
=========================================
update
Below is the folder structure of deployed codes. (Cloud functions can't show more than 50 files so I downloaded the source code from GCP)
As you can see the ormconfig.json does exist in the root, but somehow it cannot be located. I have to create connection manually with typeorm.createConnection({type: "postgres",...}) to make the code work.
This is likely being caused by a known bug with app-root-path (which TypeORM uses for config file resolution) when used in conjunction with Google Cloud Functions.
The workaround / fix that worked for me was to set the environment variable APP_ROOT_PATH to /workspace when I deployed my Google Cloud Function (app-root-path will short-circuit when it sees that variable).
Related
I want to debug my react app locally in the browser,
but i set devtool property to hidden-source-map in webpack.config.js file in order to hide my source code in production.
Is there any possibility to debug locally without exposing my source map to production?
You can enable source-map on production, but host it in a server that only allows request from whitelisted IPs. If a request came from a non-whitelisted IP, you can just return error 403 or 404.
Let's say you are uploading your source maps to amazon s3:
// we use webpack.SourceMapDevToolPlugin for more flexible setups. Set the 'devtool' option to 'false' when you are doing this.
plugins: [
new webpack.SourceMapDevToolPlugin({
filename: '[name].[contenthash].js.map',
// this is a s3 private bucket that is only accessible via whitelisted IPs
// regular user will not be able to access the bucket
append: `\n//# sourceMappingURL=https://s3.ap-southeast-1.amazonaws.com/sources-maps/[url]`,
...options,
}),
// other plugins.
]
Your generated JS will still have the magic comment at the end of the file:
//# sourceMappingURL=https://s3.ap-southeast-1.amazonaws.com/sources-maps/main.js.map
But browser devtools can only download the source map if it is accessing it from a whitelisted IP, such as your office network, a company VPN, etc.
Another approach is you can just set the sourceMappingURL to localhost. With this approach, you should have all the *.map files available locally on your machine. When you want to debug production code, simply start a static server (e. g.: ecstatic) locally to serve the source maps. This way, you can be sure only you can access the source maps. But, it requires manual work to download and serve the source maps locally.
You can use the environment variable for this. It would be something like this:
devtool: process.env.SOURCE_MAP ? 'inline-source-map' : 'hidden-source-map',
Then you could run tests like this for example:
SOURCE_MAP=true yarn test (just an example)
You should checkout this module https://www.npmjs.com/package/dotenv. It's possible to create .env files and use the configuration for different flows, like test and build, so you don't have to specify env variables manually, before running the command.
If you don't want to use the environment variables, you could create separate webpack config files, maybe one that has all the common stuff between production and test environments and then more specific configs that extend the common config (one for test env with source maps enabled and one for production with source maps disabled).
Using ReactJS I made a Build (reactJs static, npm build) and uploaded it to Google Cloud Storage Bucket, but getting a issue with the Path and Build folder files. The app (/static website) running but could not fetch the files from the bucket directory for eg the index.html & logo. (404 or 403 error )
Structure: Parent Bucket > Build folder (index.html, static folder & other files inside Build)
Any one have any suggestion on this. How to resolve this?
Do I need to create an app.yaml for GCS Bucket or any alternative?
I have gone through the article quite similar but for AppEngine instead of Bucket. https://medium.com/google-cloud/how-to-deploy-a-static-react-site-to-google-cloud-platform-55ff0bd0f509.
I have tried with app.yaml file but does not work for me.
I had exactly the same issue as mentioned by the OP. I am sharing my version of solution just in case anyone else ends up here.
As shown in the screenshots by OP, the 403 errors showed up for me because the URL of the static files in build/static folder was not correctly configured by the react-scripts build script.
Eg:
The url for index.html file was https://storage.googleapis.com/{bucket-name}/index.html.
However, when the page loaded, it requested files having url https://storage.googleapis.com/static/js/main.f555c3e0.chunk.js. It should rather be
https://storage.googleapis.com/{bucket-name}/static/js/main.f555c3e0.chunk.js
This is happening because by default react-scripts build assumes that your files will be served from root folder of your host.
To fix this add the following field in package.json of your project
"homepage": "https://storage.googleapis.com/{bucket-name}"
This tells the build script that it needs to add a relative path for serving static files.
For details please refer: https://create-react-app.dev/docs/deployment/#building-for-relative-paths
In order to set the routes of a static website stored in Google Cloud Storage, you need to assign a suffix to your objects. In other words, using suffixes is the intended way to configure your website. You can see more information in Hosting a static website document.
For your main index page you should set MainPageSuffix and for the not found page 404.html you should set NotFoundPage as suffix.
You can see more information on how to configure your static web here
I am exploring GAE with nconf and I'm wondering if the following setup is secured after I deploy an App.
What concerns me is are both my "config.dev.json" and "config.prod.json" files deployed despite including them in ".gitignore".
I am unsure what information is passed along to gae (I don't want my config keys exposed) after I do:
$ git add .
$ git commit -m 'Commiting'
$ glcoud app deploy
My Node App structure looks like this:
- /myProject
- /node_modules
- .gitignore
- app.js
- app.yaml
- config.js
- keys.dev.json
- keys.prod.json
- package-lock.json
- package.json
// .gitignore
node_modules
keys.dev.json
keys.prod.json
// config.js:
const nconf = require("nconf");
nconf.argv().env();
if (nconf.get("NODE_ENV") === "production") {
nconf.file("keys.prod.json");
} else {
nconf.file("keys.dev.json");
}
...
Including files in .gitignore has no implications whatsoever on deployment on GAE, that file is only used by git.
If you want to prevent deployment of a file to GAE you need to use the skip_files option in your app.yaml file's General settings:
skip_files
Optional. The skip_files element specifies which files in the
application directory are not to be uploaded to App Engine. The value
is either a regular expression, or a list of regular expressions. Any
filename that matches any of the regular expressions is omitted from
the list of files to upload when the application is uploaded.
For example, to skip files whose names end in .bak, add a
skip_files section like the following:
skip_files:
- ^(.*/)?\.bak$
Side notes:
if I understand correctly, your app uses those files, so it appears to me like you will have to deploy them together with your app.
even if a file is deployed on GAE it is your app's responsability (and complete control) in deciding if the file is exposed to ouside requests or not.
if you want to know exactly which files are included in the deployment you can see them displayed during deployment by using the --verbosity option for the gcloud app deploy command.
I'm building a Google App Engine application with a Go backend + Polymer frontend. As a result, I'm using a dispatch.yaml file to serve both at the same time.
The problem I'm facing is that the datastore is empty when I restart my computer. I've tested this on both OSX 10.9.5 and 10.10.4. Both exhibit the same response upon a system reboot. Windows 7, however, seems to hold on to the data.
The documentation suggests that data should persist, since I'm not explicitly calling a clear. It's not. I've tried to set the datastore location myself using this:
dev_appserver.py --datastore_path=~/go_apps/data ~/go_apps/my_app
I'm receiving this error:
google.appengine.tools.devappserver2.errors.AppConfigNotFoundError: "/Users/anthony/go_apps/my_app is a directory but does not contain app.yaml or app.yml
Obviously, since I'm using a dispatch.yaml file, it wouldn't. So, since the backend, which handles the data, does have an app.yaml file, I try to set it there. I use this command:
dev_appserver.py --datastore_path=~/go_apps/data ~/go_apps/my_app/backend
That doesn't seem to work either, as I get this error:
sqlite3.OperationalError: unable to open database file
Okay? Well, not sure where to turn now. From what I could gather from other posts, that data is stored temporarily. But, I can't seem to set a custom, non-temporary location for the data. So, now I'm populating a datastore every time I reboot, which seems ridiculous.
* Edit *
I've tried the following, which seems like it tries to launch the app, and creates a datastore.db file at the correct location:
dev_appserver.py --datastore_path ~/go_apps/my_app/data/datastore.db ~/go_apps/my_app/dispatch.yaml ~/go_apps/my_app/backend/app.yaml ~/go_apps/my_app/frontend/app.yaml
However, I'm getting a weird error now:
/var/folders/04/3hxnpxc15wj2k4v40lkdncd00000gn/T/tmpkcQYnFappengine-go-bin/backend.go:13: can't find import: "github.com/gorilla/mux"
Does Go build to that folder temporarily? That import is definitely available, and always builds fine calling goapp serve.
Here is what my imports look like on backend.go
import (
//standard library
"fmt"
"net/http"
"time"
"log"
//third party
"github.com/gorilla/mux"
"github.com/gorilla/securecookie"
"github.com/dgrijalva/jwt-go"
"golang.org/x/crypto/bcrypt"
//my imports
"github.com/section14/go_polymer_comm_pkg/controller"
)
You have to pass the name of the file to be used as the persisted datastore, not a folder.
And next provide the folder of your app (which contains app.yaml). Don't mix the 2. So it should be something like:
dev_appserver.py --datastore_path=~/my_app/my_app.db ~/go_apps/my_app
Details can be found here:
The Go Development Server / Using the Datastore
Notes:
The default datastore file is in the temp folder, and your OS-X most likely clears that on system restart, that's why it is not preserved for you. On the other hand Windows 7 for example does not clear the temp folder on system restarts.
Got it up and running by adding both GOPATH and GOROOT environment variables to my .bash_profile. In total, these three paths (first path was already set) are needed for it to run:
# Add Google AppEngine path
export PATH=/Users/anthony/go_appengine:$PATH
# GOPATH
export GOPATH=/Users/anthony/go_appengine/gopath
export PATH=$PATH:$GOPATH
# GOROOT
export GOROOT=/Users/anthony/go_appengine/goroot
export PATH=$PATH:$GOROOT
This command is called from inside the project folder (mine resides outside of the appengine folder) for it to launch:
dev_appserver.py --datastore_path data/datastore.db dispatch.yaml backend/app.yaml frontend/app.yaml
Notice that the .yaml files are still there. It builds fine with them, and probably builds fine without them if you don't need a dispatch.yaml file.
Thanks #icza for the direction. Wanted to organize the steps in a post for easier reading.
similar questions have been asked before, regarding grails 2(.3, .4). I find it strange that i could not find a way to do this, as it seems a standard use-case to me.
I simply want to serve html-pages, including their linked .css and .js (angular and jquery content) when i run grails run-app.
I want to check if my http-calls are handeled correctly on both sides - without needing to deploy a .war and configuring a database.
afaik grails run-app simply starts a jetty/tomcat - both of which can obviously serve .html pages. What do i have to do to make the grails development-tooling deploy my files?
I need to make http-requests,
so using a different Server would violate JS-SOP,
while deploying the .war would greatly slow down the development process
I've so far only found clunky jsonp, proxy, .war deployment solutions, or solutions for grails 2.x
I tried placing the files literally everywhere in the projects' structure (/src/main, /src/main/resources, /src/main/public, the assets folder and its subfolders, created web-app directories in every subdirectory, the Init, domain, conf directories - you name it)
Add the index.html to src/main/resources/public
Then add this to UrlMappings.groovy:
"/"(redirect:"/index.html")
For grails >= 3.0.12
Location of static resources
In order to resolve an issue around how POST requests are treated for
REST applications on non-existent resources, static resources located
in src/main/resources/public are now resolved under the /static/** URI
by default, instead of the base URI /**. If you wish to restore the
previous behaviour add the following configuration:
grails.resources.pattern = '/**'
https://github.com/grails/grails-core/releases/tag/v3.0.12
Contrary to the accepted answer, you don't need a redirect. I have made able to make this work with the following config:
UrlMappings.groovy
"/"(uri: "/index.html")
application.yml
grails:
resources:
pattern: '/**'
Finally, you just need to have your index.html file located under src/main/webapp