-
Notifications
You must be signed in to change notification settings - Fork 5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Closing #30 by connecting to Heroku #53
Conversation
…heck travis behavior
… use the homepage url
…removing our build folder from .gitignore so we can easily push up the build.
A couple questions:
|
@phuchle
Heroku Dashboard supports auto-deployments for specific branches and we can set up pipelines there. I know the auto-deployments can build the pages we want but won't do "react-scripts build" since it only installs dependencies and not DevDepends. I think to achieve this you would want to use the Pipeline feature which I haven't deep-dived yet.
I can write one if we need it, but it would probably be mostly around Heroku since the only thing we have to do in AAF is locally run |
OK, I think that can be saved until later on in the dev. pipeline.
Yeah that would be great. Can we review the steps on discord whenever you're free so I can learn how the deployment goes down? |
@phuchle Yeah I think for now we can keep the build files but we definitely want to revisit later. However, in the end when we host this, we would be pushing build files and storing them on the server. I really do think that setting up the pipeline process has to be done eventually. I can show everyone how to use the Heroku environment. If any of the team members don't have access to Heroku I can add them later too. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
To the extent of my backend knowledge, this looks good
If heroku reads from the designated branch build folder, could we just make sure our build script sets up everything we need, if not, would we need to consider looking into a simple task runner like gulp? |
@jwu910 I think webpack handles the compiling and transpiling. All heroku does (right now) is read the build folder that react-scripts creates. Eventually we should sophisticate our pipeline to include more than 1 step. We can discuss this over discord in more detail on tuesday |
* Uncommenting logic to use build folder. Also using double quotes to check travis behavior * Adding `build` to source control * Adding Procfile to have heroku run a specific server command * Mofidying the path for express public folder * Modyfing homepage field on package.json and pushing up new build that use the homepage url * Adding build folder to eslint ignore file so we do not fail CI. Also removing our build folder from .gitignore so we can easily push up the build.
* Uncommenting logic to use build folder. Also using double quotes to check travis behavior * Adding `build` to source control * Adding Procfile to have heroku run a specific server command * Mofidying the path for express public folder * Modyfing homepage field on package.json and pushing up new build that use the homepage url * Adding build folder to eslint ignore file so we do not fail CI. Also removing our build folder from .gitignore so we can easily push up the build.
* Uncommenting logic to use build folder. Also using double quotes to check travis behavior * Adding `build` to source control * Adding Procfile to have heroku run a specific server command * Mofidying the path for express public folder * Modyfing homepage field on package.json and pushing up new build that use the homepage url * Adding build folder to eslint ignore file so we do not fail CI. Also removing our build folder from .gitignore so we can easily push up the build.
* Uncommenting logic to use build folder. Also using double quotes to check travis behavior * Adding `build` to source control * Adding Procfile to have heroku run a specific server command * Mofidying the path for express public folder * Modyfing homepage field on package.json and pushing up new build that use the homepage url * Adding build folder to eslint ignore file so we do not fail CI. Also removing our build folder from .gitignore so we can easily push up the build.
* Closing #30 by connecting to Heroku (#53) * Uncommenting logic to use build folder. Also using double quotes to check travis behavior * Adding `build` to source control * Adding Procfile to have heroku run a specific server command * Mofidying the path for express public folder * Modyfing homepage field on package.json and pushing up new build that use the homepage url * Adding build folder to eslint ignore file so we do not fail CI. Also removing our build folder from .gitignore so we can easily push up the build. * Install mongoose and bcrypt for writing models. * Add base user model and placeholder files for donor, family, and organizer models. * Add more fields and refine methods on userSchema. * Run user.js through prettier. * Add wishlist schema. * Refine existing models for consistency and clarity. Add initial version of other base models. * Add a few more methods before realizing I have no idea what I'm doing. * Update models. Add donor and family controllers. Add body-parser package for dealing with form data. * Setting up a very basic set of api routes * Adding basic logic for /pairing route. Modifying Donor and Family schema to make references to each otehr for matching and adding default values to it. /pairing should find all donors and filter them based on budget then sort budgets in descending order. * Adding organizer to the donor and family models * Adding pseudo logic for /paring/balance route * Adding notes to /pairing/balance to remind us to use the readWishlist() method * Correcting issues with api.js Switching the min/max budgets in the LTE/GTE query and removing the p from "res" * SeedDB script is working now (#93) * Fix capitalization error in server/models. * Update ecmaVersion to latest in order to use async/await. * Implement working seedDB script by making liberal use of async/await. * Add consistent error/sucess output on saving each model. * Aaf 97 budget pairing (#101) * Fixing #100 - Removing extra key, field, from Donor.sort * Pulling the data from the request correctly so we can find budgets accordingly * Fixing #97. The max budget cap is not implemented so we sort at a +/- 10% in ascending order for now. Also did some work for setting the database route to 'aaf' in our api file. Also setting matchedFamily to null in our models and correcting the name mapping in the db seed file. * Correcting the sort order to be as #97 is asking. * Add wishlist seeding to seedDB script (#102) Add wishlist seeding to seed DB and refine error handling for wishlists and families. * AAF-104 Change body parser declaration from app to routes * Api test setup (#106) * WIP :: Setting up API tests with base skeleton * Modifying app.js to export the app.listen so that the tests can run correctly. Setting up a single happy path test for /api * Enhancing api tests to make proper assertions. * Adding logic to /api/pairing to handle missing budget params. Making tests more relevant to pairing route. * Adding tests for making sure messages are only should when the appropriate number of donors are returned * Removing duplicated bodyparser line in api.js * Cleaning up the Pairing test to split the tests into sub-sets. Also adding some basic wishlist test templates for the wishlist routes. * Removing the babel dependencies that were added since they aren't actually used in testing. * Fixing the issue where only one test would run. Also, since multiple tests can no run you can watch the api tests using $(npm bin)/jest api --watch This assumes you're naming api tests as <ROUTE>Api.test.js * Removing useless comment on app.js * Fixing the linter issue for wishlist route so npm test will execute * Fixing the unused imports issue * AAF-111 Add usePushEach as deprecation workaround for mongo that affects array push methods (#112) * AAF-111 Add usePushEach as deprecation workaround for mongo that affects array push methods * AAF-111 Npm update and package lock * AAF-111 Add jest-cli as dev dependency * AAF-94 wishlist api read (#108) * add first route for families to read their wishlist * Add more error handling. Add a family with an empty wishlist. * AAF94 Add a description for the regex Fix the logic for the if statement Change the response status to 400 and made the output messages more discriptive * AAF 94 Change the HTTP code for no family found to 404. * AAF-94 Fix uppercase for const variable * AAF-94 Revert the change to the familyId variable and fix the route * AAF-94 Remove the delete route * AAF-113 npm test api -- Fix #113 (#114) * Trying to fix #113 by using concurrently to run all the tests but I doubt that this will work correctly. I might have to make the changes at the travis.yml file * (#113)Telling travis to cache node_modules so that it doesn't have to constantly install if things don't change * 113 - Skipping the wishlist tests since they aren't actually written. However, when travis runs concurrently, it might be failing to make a second connection to mongo.. will need to test this * AAF#113 - Adding mongodb service and seed script to travis.yml to solve the issue with scripts not running. * AAF-#113 - Skipping the POST routes tests in the wishlist API * Removing node 4 and 6 from Travis since we won't be supporting that in prod. Also modifying our README to show system requirements that are recommending a minium node version of 7 and a mongodb version of 3.6 (#113) * Modifying the test:api command to now run on a node nevironment since we're testing API (node services) and not some dom elements. Also setting API test to trigger OS Notification and to force quit once the execution is done ( no watch mode ) (#113) * moved the totalListCost outside of the list array (#117) * Aaf 109 families filter (#122) AAF-109 Add families filter route. This route is expected to be used like so: - GET /api/families with no queries returns list of all families - GET /api/families?filter=_id&value=123 Passing filter and value will look up families with the provided filter and value and return an object with a property families, an array of the families. If you pass less than 2 or more than 2 query fields, the route throws 400. If you pass exactly 2 query fields but they are not filter or value, it will throw 400. eg: filter: foo and valu: bar or foo: bar and bar: foo * Aaf 110 donors filter (#126) * AAF-109 Write show and filter route for donor. Experimenting with controller abstraction. * AAF-110 Rebuild bcrypt and node-sass due to errors that caused npm start to break. * Removing bcrypt from our package.json since we do not need this package... i dont think * readding bcrypt as version 2.0.1 since we are using it in seed file. * Setting up Bcrypt v3.0 * AAF-63 Update node-sass-chokidar. 1.3.0 released. node-sass dependencies should build with node-gyp * AAF-63 Autogenerate packagelock
No description provided.