Wasp-lang•3d ago

Update / maintain production deployment

Lets say i deploy to fly my V1 of my app. How would i then maintain it and for example update it without having users loose data. Let say i add a new page and some new actions locally. How can i push these to fly?

42 Replies

kapa.ai•3d ago

Hi @Sven. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave: ---------------------- To update your deployed app on Fly with the latest changes, you can use the wasp deploy command. Here's how you can do it:

wasp deploy fly deploy

wasp deploy fly deploy

This command pushes your client and server live. Run this command whenever you want to update your deployed app with the latest changes [^1^]. If you are deploying manually, for example on Railway, you would need to run wasp build to rebuild your app and then railway up in the .wasp/build directory (server). Repeat all the steps in the .wasp/build/web-app directory (client) [^3^]. When it comes to data migrations, it's important to have a good migration workflow and documentation for production. This is especially crucial when you're updating your app to ensure users don't lose data [^9^]. If you need to modify the server code a bit to execute some code after login / after signup, you can modify the .wasp/out files. However, when deploying to Fly, Wasp builds a fresh build in .wasp/build so those manual changes won't be deployed. One way to solve this is to have a "patched" version of the file you need to modify, somewhere in the project root. Then you provide your own custom Dockerfile (which can be just the copy-paste version of the Dockerfile in .wasp/build folder). In that Dockerfile include a line with something like cp ./my-patched-file.js server/something/somewhere.js [^2^]. [^1^]: (source: https://wasp-lang.dev/docs/advanced/deployment/cli#deploy) [^2^]: (source: https://discord.com/channels/686873244791210014/1223671727943319584) [^3^]: (source: https://wasp-lang.dev/docs/advanced/deployment/manually#updates--redeploying) [^9^]: (source: https://github.com/wasp-lang/wasp/issues/330) ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Sven, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?

Sven•3d ago

Huh?

Adam•3d ago

A migration doesn't break your production database Use wasp deploy fly deploy to deploy to prod

Sven•3d ago

Ahh so then it updates all stuff And lets say i add an env var i use -server-secret? Or secret set Thanks to

Adam•3d ago

It's in the wasp deploy docs but yeah something like that

wardbox•3d ago

This is not in all cases, it depends what you’re changing. Some migrations can require manual intervention

Sven•3d ago

Hmm why is that?

wardbox•3d ago

Let’s say you have a field that’s a string in prod, if you’re changing it to a Date field you’re going to run into issues If you do migrate dev locally and it runs into issues like that, the same thing will happen in prod. You can ssh to the db and fix it in that case though

MEE6•3d ago

Wohooo @wardbox, you just became a Waspeteer level 9!

wardbox•3d ago

Usually when I need to change a field like that i add a brand new field with a slightly different name, leave the old one and make it optional if it isn’t already, and update references to the field to use the new one with the new type. Then once all references to the old field are eventually not in use you can just remove the old field from the schema But if you aren’t changing types or adding unique constraints, most of the time the migration is fine

Sven•3d ago

Yeah thanks clear

Adam•3d ago

Sure - I agree, any migration that requires manual intervention on your local development side as well. So when wasp db-migrate dev requires manual intervention then manual intervention in prod -- if you or anyone from wasp could make a slight guide or tell me some tips about how to do this this would be great. This is the only part which I'm queasy about. The SQL migrations when new fields are added - do you have any helpful tips/links about that?

MEE6•3d ago

Wohooo @Adam, you just became a Waspeteer level 4!

wardbox•3d ago

https://www.prisma.io/dataguide/types/relational/migration-strategies not wasp specific, but this outlines strategies for handling migrations

Prisma's Data Guide

Migrating database schema changes | Prisma's Data Guide

Learn about the advantages and disadvantages of various techniques for deploying schema changes to your database.

Adam•3d ago

Awesome thanks!

Sven•2d ago

But ward, Lets say I changed somethings on the frontend. And added some actions. How would i makensure these changes get to production? After i already deployed the first version?

wardbox•2d ago

You just do a new deploy

Sven•2d ago

any tips on connecting to prod. database with ext tools? to view tables

wardbox•2d ago

@kapa.ai how do you connect over wireguard to fly.io Postgres database with wasp db studio

kapa.ai•2d ago

Let me check the knowledge base for you again. One moment! :books: ---------------------- To connect to your Fly production database, you can open up a tunnel with the following command:

fly proxy 5432 -a <db-name>

fly proxy 5432 -a <db-name>

After setting up the tunnel, you can execute wasp db studio to interact with your database using Prisma Studio [^1^]. However, please note that you should be careful when interacting with your production database. It's recommended to remove the DATABASE_URL with the production database URL from .env.server after you're done, to avoid accidentally making unwanted changes to the production database [^1^]. For more detailed steps on how to seed a production database, you can refer to this Github issue (https://github.com/wasp-lang/wasp/issues/1464) [^5^]. [^1^]: (source: https://discord.com/channels/686873244791210014/1220423321422663820) [^5^]: (source: https://github.com/wasp-lang/wasp/issues/1464) ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @wardbox, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?

Sven•2d ago

@kapa.ai how i ran the fly proxy 5432 -a <db-name> , how can i now show the prisma studio in browser?

kapa.ai•2d ago

Let me check the knowledge base for you again. One moment! :books: ---------------------- After you've set up the tunnel with fly proxy 5432 -a <db-name>, you can run Prisma Studio with the following command:

wasp db studio

wasp db studio

This command will start Prisma Studio and automatically open it in your default web browser [^1^]. Please ensure that you've set the correct DATABASE_URL in your .env.server file. If Prisma Studio fails to connect, it might be due to an incorrect DATABASE_URL (wrong password or database name), or something else might already be running on port 5432 [^5^]. [^1^]: (source: https://discord.com/channels/686873244791210014/1220423321422663820) [^5^]: (source: https://github.com/wasp-lang/wasp/issues/1464) ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Sven, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?

Sven•2d ago

ah cool clear. now i also have my own server. What I want it for it to fill the production database table "datatable" with data everyday (python). I already have it working with a different postgres db. but how would i connect to it using my production DB? psycopg2 is what i use.

wardbox•2d ago

Not sure, I haven’t used python with wasp

Sven•2d ago

ahh but like how could i connect from an external server? what credentials do i use? @kapa.ai I have an app running on a server, how can iconnect to my production database (fly) through this python app? I want the app to inset data into a table everyday/

kapa.ai•2d ago

Let me check the knowledge base for you again. One moment! :books: ---------------------- To connect to your production database from a Python app, you'll need to use a PostgreSQL client for Python, such as psycopg2 or sqlalchemy. You'll also need the connection details for your database, which include the host, port, database name, username, and password. Here's a basic example of how you might establish a connection using psycopg2:

import psycopg2

# Replace these with your actual values
host = "localhost"
port = "5432"
dbname = "<db_name>"
user = "postgres"
password = "<password>"

# Establish a connection
conn = psycopg2.connect(
    host=host,
    port=port,
    dbname=dbname,
    user=user,
    password=password
)

# Create a cursor object
cur = conn.cursor()

# Execute a query
cur.execute("SELECT * FROM your_table")

# Fetch all rows from the last executed statement
rows = cur.fetchall()

# Close the connection
conn.close()

import psycopg2

# Replace these with your actual values
host = "localhost"
port = "5432"
dbname = "<db_name>"
user = "postgres"
password = "<password>"

# Establish a connection
conn = psycopg2.connect(
    host=host,
    port=port,
    dbname=dbname,
    user=user,
    password=password
)

# Create a cursor object
cur = conn.cursor()

# Execute a query
cur.execute("SELECT * FROM your_table")

# Fetch all rows from the last executed statement
rows = cur.fetchall()

# Close the connection
conn.close()

To get the connection details, you can use the fly postgres connect -a <db-name> command to get the DB name [^5^]. For the DB password, you can use the fly ssh console -a <db-name> command and then echo $OPERATOR_PASSWORD [^5^]. Please note that this is a basic example and doesn't include error handling or other best practices for working with databases in Python. You'll also need to install the psycopg2 package, which you can do with pip install psycopg2. For inserting data into a table every day, you might want to look into using a task scheduler like cron on Unix-based systems, or Task Scheduler on Windows. [^5^]: (source: https://github.com/wasp-lang/wasp/issues/1464) ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Sven, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins? Let me check the knowledge base for you again. One moment! :books: Sorry @Sven, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?

Sven•2d ago

but this is for localhost, how can i conect from a app running on a server? @kapa.ai

kapa.ai•2d ago

Let me check the knowledge base for you again. One moment! :books: ---------------------- To connect to your Fly production database from a Python app running on a server, you'll need to replace the localhost in the connection string with the actual host of your Fly database. The host is typically a URL or IP address that points to the location of your database on the internet. You can find this information in the settings or configuration of your Fly database. Here's how you might modify the previous Python code:

import psycopg2

# Replace these with your actual values
host = "<your-fly-db-host>"
port = "5432"
dbname = "<db_name>"
user = "postgres"
password = "<password>"

# Establish a connection
conn = psycopg2.connect(
    host=host,
    port=port,
    dbname=dbname,
    user=user,
    password=password
)

# Create a cursor object
cur = conn.cursor()

# Execute a query
cur.execute("SELECT * FROM your_table")

# Fetch all rows from the last executed statement
rows = cur.fetchall()

# Close the connection
conn.close()

import psycopg2

# Replace these with your actual values
host = "<your-fly-db-host>"
port = "5432"
dbname = "<db_name>"
user = "postgres"
password = "<password>"

# Establish a connection
conn = psycopg2.connect(
    host=host,
    port=port,
    dbname=dbname,
    user=user,
    password=password
)

# Create a cursor object
cur = conn.cursor()

# Execute a query
cur.execute("SELECT * FROM your_table")

# Fetch all rows from the last executed statement
rows = cur.fetchall()

# Close the connection
conn.close()

Remember to replace <your-fly-db-host>, <db_name>, and <password> with your actual database host, database name, and password respectively. Please note that exposing your database to the public internet can have security implications. It's recommended to secure your database connection, for example by using SSL/TLS encryption or by setting up a VPN or SSH tunnel. You might also want to restrict access to your database to specific IP addresses or ranges. For more information on how to connect to your Fly database, you can refer to the Fly documentation (https://fly.io/docs/) or the Fly CLI command reference (https://fly.io/docs/flyctl/). ---------------------- Note: you can ask me a follow up question by @ mentioning me again :speech_balloon: ---------------------- Sorry @Sven, I seem to be having a bit of trouble. Don't worry though, I've pinged my support and they will try to fix me up! :technologist: Perhaps try again in 15mins?

Sven•2d ago

any wasp team that could help, i would like to connect to my fly.io database from a python app running on a server. which does some actoins everyday and inserts data in a table. Is there any way of connecting to my production fly postgres (from a server)? Any tips? Maybe should i deploy my database somewhere else? And how would I make my app connect to that one then? Just secret set “DATABASE_URL” to my other host? And will this then automaticallly create the tables? @Wasp Team I know u guys don like me tagging, but i have been searching the whole day and nothing works...

Adam•2d ago

Not sure why you need to manually add data in your production app, but did you try seeding it? They have information about seeding. If you need to access your production database, when you deployed to fly you needed to press a button after they showed you the database url and username and password, download dbbeaver and use that to connect and modify if you need. Yeah I think we'd all benefit from a database migration document I'll work on writing something as well - I actually have been running into snags as well, for example I change something in my schema.prisma, i push to prod and expect the GitHub action to run the migration, but it doesn't pick the migration up. I've had to delete my app to pick up the migration, which is curious - I've looked at my logs and see there are conflicts, so maybe it is just my problem

Sven•2d ago

Hmm well, my app provides valuablw information about searchdata for the past 30 days. So that table needst to be updated everyday. I already have a working py script that updates these in a test table. Now i want to connect to my production table from thst “python app on my server”

Adam•this hour

Interesting approach, you have a local db that you want to connect with the prod database. I don't think that's really wasp specific, I would look into how to connect my database to the fly database. Have you used cursor? It gives really good help You also could have a job that does specific actions on your server, modifications of your database and so. I would probably do that approach actually, to forgo some weird joins, but your mileage may vary. Definitley try out cursor

Sven•this hour

well i already have a working script i want to use as it inovlves web scraping. And i just want my python app running on heroky to be able to connect to my fly postgres to insert data

MEE6•this hour

Wohooo @Sven, you just became a Waspeteer level 11!

Adam•this hour

ok, yea I see would be great if wasp wanted to help pro bono, but I that's more app specific I think

Sven•22h ago

so yeah any help would be apreciated Okay Solved!

Sven•22h ago

but i follwed these steps: https://fly.io/docs/postgres/connecting/connecting-external/

Fly

External Connections

Documentation and guides from the team at Fly.io.

martinsos•15h ago

Yup this is the best way to do it, I tihnk there is even a name for it, forward migrations or something like that, but it is the best practice.

martinsos•15h ago

@Adam you are right, we could do with better docs on migrations for sure, and also maybe a bit better setup on Wasp side (gihtub action, even in docker container). We have an issue open for it here, hasn't been high priroty so far, but will happen before 1.0 for sure: https://github.com/wasp-lang/wasp/issues/330 .

GitHub

Support data migrations & provide best practices and examples · Iss...

While developing Waspello, at some point I wanted to migrate data in my db. I added pos field to List entity so I wanted to assign a default pos value to each existing List, otherwise the changed s...

martinsos•15h ago

@Sven -> this, what @Adam said. Best approach here would most likely be to make this a job in your Wasp app. This is safer, and should be easier to manage because python script now ships with your wasp app, you can version them together, you can test it locally, ... . You could rewrite that python to JS and do it that way. Or you can even keep using python script, but call it from wasp's job, from JS, get the result from your python script, and then have JS update the database with it, via Prisma. So python does the scraping, JS does putting into the database. That way you can utilize Prisma. How can you call python from JS? Easiest way is probably to just run it as a process on the machine (something like child_process.execSync) and get output from its stdout, or have it write it to a tmp file and then read it from there. Here you can find more info on how to possibly run python from Wasp https://github.com/wasp-lang/wasp/issues/1940 + you can see what some other users did.

Sven•15h ago

Okay cool, hmm i wold like to keep them seperated because the scraping can be intensive etc. overall it should not matter as it just appends data daily and that data is used in my dashboard

martinsos•15h ago

If scraping is intensive, than that is a good argument to keep them separated, that is true. Is it really so intensive that you need another server machine though? You will know that the best!

Gaming

Programming

Update / maintain production deployment