if you need subscriptions, you should move to something like hasura, that was the advice my team got from a prisma engineer

And migrations, updating the DB is all easy?

and we just do a quick rewrite and hook that up to our graphql layer that is currently in node

We have Postgress > Prisma > Graphql > Web App + Mobile App

if you use join tables, moving to anything will be pretty hard, so hopefully you only use inline connections

yes we have not used any links

hasura would connect to postgres and basically act as your backend as they remove the need for a node backend in most cases.

you would protect your data with claims in your jwt

so essentially we would remove the graphql layer

no, sorry, hasura is graphql.

Hi, guys. I have an issue with datamodel.prisma (I think):

enum Permission {

ADMIN

USER

ITEMCREATE

ITEMUPDATE

ITEMDELETE

PERMISSIONUPDATE

}

type User {

id: ID! @id

name: String!

email: String! @unique

password: String!

resetToken: String

resetTokenExpiry: Float

createdAt: DateTime! @createdAt

updatedAt: DateTime! @updatedAt

permissions: [Permission!]! @scalarList(strategy: RELATION)

}

and then I create a mutation with query

ctx.db.query.user({ where: { email } });

but get return without

permissions

field. "message": "Cannot return null for non-nullable field User.permissions." On prisma.io server, User have this field with a set value

Hi guys, prisma 2 doesn’t have docker image? i wanna make my backend project with docket and prisma 2. but prisma 2 image doesn’t exit in docker hub

Hi, I haven't found info about it in the docs: • Is it possible to change a scalar type field to an array of the same type, without losing the data?

When will Prisma v2 have MongoDB support?

Is anybody can help me? https://prisma.slack.com/archives/CA491RJH0/p1587035647124300

Now, I have 1 database with Prisma. Models: `User`/`Post`/`Log` The database is starting to be huge, because of the table

Log

. What would be the Prisma architecture to have 2 distincts databases (`User`/`Post` and

Log

)

Hi there, Is there any information available on how to create a relationship on oneself with Prisma 2?

Is there a way to create the entities relation tree ?

Does anyone have instruction on how to backup and restore prisma PG database? I could never get it right and prisma always return errors Thx!

Does anyone have any smart workarounds for json fields when upgrading a prisma1 project?

Hi, I have a question about nesting includes and selects. As an example, I have Table A with several feeds including a relation field that connects it to Table B. If I say

...
				return ctx.prisma.TableA
					.findMany({
						where: {
							...
						},
						include: {
							TableBconnectionField: true					},
					})
...

I get all the fields in Table A and all the fields in Table B. But I also have a connection field in Table B to Table C and I need some data from that included in the return JSON as well. If I put

...
				return ctx.prisma.TableA
					.findMany({
						where: {
							...
						},
						include: {
							TableBconnectionField: {
								select: { field_1: true, field_2: true, TableCconnectionField: true },
							}
						},
					})
...

I get all the fields for Tables A and C but only get field_1 and field_2 for Table B. Is there a way to do this that gives me all the fields for Table B as well without me having to explicitly request them?

Hey everyone, I'm trying to display a many to many relation (I guess it's one to many in this one) for a GET request. Is there a better way of doing it other than searching the db twice? How would I chain the .resources() call to the first function?

app.get(`/projects/:id`, async (req, res) => {
   const { id } = req.params;
   const project = await prisma.project.findOne({
      where: {
         id: Number(id),
      },
   });
   const resourcesOfProject = await prisma.project
      .findOne({ where: { id: Number(id) } })
      .resources();
   res.status(200).json({ ...project, resources: resourcesOfProject });
});

// Here is some of my schema
model Resource {
  id          Int       @default(autoincrement()) @id
  name        String
  description String?
  projects    Project[] @relation(references: [id])
}

model Project {
  id          Int        @default(autoincrement()) @id
  name        String
  description String?
  tasks       Task[]
  resources   Resource[] @relation(references: [id])
  completed   Boolean    @default(false)
}

What is the best way to fill the new column for existing rows? The migration CLI is able to do this?

Use @default and then remove @default after up?

Hi all. Great to be here. 😀 I'm wondering, what might be the best way to change the datasource URL in production on Google Cloud?

Actually, I'm going nuts. What is the alternative to using an .env file when the schema.prisma file isn't JS?

What do you suggest for security when using raw queries against SQL injection? I am looking for some escape function. If possible, a solution from prisma would be the best, otherwise from another dependency

I have a question about chaining include or select statements. I want to make query that pulls a list of items and gets data from 3 or more levels. For example, I have this query:

return ctx.prisma.lead
					.findOne({
						where: args,
						include: {
							User: true,
							LeadCampaign: { include: { LeadSource: true } },
						},
					})

It successfully pulls the right row from the lead table and also the associated row from the associated LeadCampaign table and the correct row for the LeadSource that's associated with it. However, when I try to get a list of leads with their associated data like so:

return ctx.prisma.lead
					.findMany({
						where: args,
						include: {
							User: true,
							LeadCampaign: { include: { LeadSource: true } },
						},
					})

I get the error:

Expected at most 1 item for 'LeadSource', got 21

I found in the documentation that "_The only requirement for chaining is that the previous function call must return only a single object (e.g. as returned by a 

findOne

 query or a "to-one relation" like 

profile.user()

)._" So I get that it can't pull a list of items from a parent that was also a list, but each lead has only one LeadCampaign and each LeadCampaign only has one LeadSource, so it seems like it's meeting the "to-one relation" part of that statement. Do I need to write it differently? Or do I need to write a function to pull different queries and stitch the results together myself?

Hi all 👋 I have a question about the docker-compose file and the prisma config. Currently I’m passing a lot of sensitive data (like db username, password, management secret) in docker-compose file. Is there a best practice on how do extract those away into a separate file or usually you guys just exclude it in .gitignore?

Hi all I want to know that can we use Prisma2 for GraphQL?(postgres) If not, could you plz tell me why it's impossible. Thanks!