Hello, I've been using prisma migrate(great funtionality!) for some time, but somehow when I typed the command just now, it doesn't work.

npx prisma migrate save --experimental

Hi. Thank you for the new [prisma migration](https://www.prisma.io/blog/prisma-migrate-preview-b5eno5g08d0b). I have a question on applying this to github workflow. We need to run

prisma migrate dev

and

prisma migrate deploy

to update our remote database and this should be done before merging the

PR

since the queries won’t work without it. The sourcecode will be deployed to production when merged to master. The problem is, when we run

prisma migrate dev

the prompt appears (Type in migration name) and we need to type in something to apply it. Should we run it before pushing the commits to

branch

that would be merged to

master

? Currently, in our opensource project (https://github.com/dooboolab/hackatalk), we are doing

prisma migrate dev

and

prisma migrate deploy

locally when we updates

schema.ts

but I am wondering if this could be done over the air.

What can we do if the

primsa migrate dev

failed?

Error: The migration 20210111021729_item_style_none failed when applied to the shadow database.
Fix the migration script and run yarn prisma migrate dev --preview-feature again.

P3006
Migration `20210111021729_item_style_none` failed to apply cleanly to a temporary database. 
Error:
Database error: Error accessing result set, column not found: itemStyle
   0: sql_migration_connector::flavour::postgres::sql_schema_from_migration_history
             at migration-engine/connectors/sql-migration-connector/src/flavour/postgres.rs:174
   1: sql_migration_connector::sql_database_migration_inferrer::validate_migrations
             at migration-engine/connectors/sql-migration-connector/src/sql_database_migration_inferrer.rs:89
   2: migration_core::api::DiagnoseMigrationHistory
             at migration-engine/core/src/api.rs:148
error Command failed with exit code 1.
info Visit <https://yarnpkg.com/en/docs/cli/run> for documentation about this command.

I am not sure if I should use migrate deploy or push

RUN npx --no-install prisma migrate deploy --preview-feature

for my Docker command 😅

It runs everytime I deploy my project

anyone ever seen this error before running

prisma migrate up --experimental

(this is in version 2.12): ‘Error: Error in command input: Could not load migration from database.’

hi, is there a way to run

prisma db seed

when prisma is in a custom folder? we're using nx to manage our monorepo and i can't find a flag to customize the path

Hello folks 👋🏻 I have a question regarding the character set for generated migrations. I need to use

utf8

and I saw this issue where that behaviour was changed: https://github.com/prisma/prisma-engines/pull/889 Is there a way to specify it or do I have to first create the migration, modify it and then run it? Thanks 😊

Hello again! 👋🏻 I have one more question, I’m trying to introduce Prisma Migrate to an existing project. I’m using nativeTypes, and there is something strange happening every time I run

prisma migrate dev --preview-feature

(even when there’s no changes). In the .sql files there are a bunch of ALTER TABLE statements modifying every TINYINT to the same thing that they already are in the database. And it’s being generated every time, looks completely the same. This is an example:

ALTER TABLE `category` MODIFY `active` TINYINT NOT NULL DEFAULT true;

I was looking out for issues, and found this: https://github.com/prisma/prisma-engines/issues/1257. I’m not sure if that’s the same issue, if not I can open a new one 🙂 Thanks!

When running

prisma migrate dev --preview-feature

with prisma 2.14.0, I get the following error:

Error: A migration failed when applied to the shadow database:
P3006
Migration `20210128090300_init` failed to apply cleanly to a temporary database. 
ERROR 42000 (1142): ALTER command denied to user

I think the problem is that the database user does not have the required access to the shadow database. Only granting

CREATE, DROP ON *.*

like stated in the docs, seems to be not sufficient.

anyone know how to trouble shoot this? only thing I could find online so far was… dropping the database & rebuilding from scratch

which feels subpar 😅

just went ahead & dropped & rebuilt

hopefully doesn’t happen again

Do you guys add native types to every field? If not, when do you not add them?

Hey guys, as I understand prisma migrate files working in change based way. I'm using github actions for cd, so everytime i want to make a change on database i have to add a one more migration file to the repo. is there a way to keep migration in a single file?

https://github.com/prisma/prisma/issues/5383 posted this issue a bit late so not sure if anything was done since I first mentioned it here, it actually seems to magically insert into the database properly when using enum values 🤔 but it won't work when comparing strings

Is

prisma/migrations/migration_lock.toml

intended to be commit to git, or does it just represent local state?

I am using digital ocean’s database to benefit of it’s connection pooling. I am trying to use the

prisma migrate reset --preview-feature -f

command to run tests against my preview environment created by vercel. However I get

Error: Database error: Error querying the database: db error: ERROR: must be owner of schema public

On DigitalOcean the

public

schema is created by the role

postgres

which is AFAIK not accessible. Is there any solution to this problem?

On a related note: Has anyone seen this? https://github.com/oguimbal/pg-mem would be awesome if it would work with prisma out of the box. Will try around with it this week

How come sometimes when applying a migration it asks to create a new one?

Hello! Sorry about the radio silence. We’ve been really busy since the start of the year. Here are a few updates of what we’ve been up to. https://github.com/prisma/prisma/issues/4531#issuecomment-771893085

Update: 2.15.0 release In 2.15.0 we added support for native types in Prisma Migrate via the

nativeTypes

preview feature. Using the native type syntax for PSL, Prisma Migrate will generate the corresponding DDL statements in the migrations. We also added support for seeding. You can find the guide here and the reference docs here. Feedback for the seeding functionality can be shared here. Prisma Migrate will automatically trigger seeding when the database is reset. More info in the release notes for 2.15.0.

Update: 2.16.0 release As you may have noticed most of our team effort in the last two weeks has been spent on native types and improving its compatibility with Prisma Migrate. We’ve now added

Unsupported(String)

and

dbgenerated(String)

, which makes it possible to use Prisma Migrate to manage columns of types that are not yet supported by Prisma. Example:

generator client {
  provider        = "prisma-client-js"
  previewFeatures = ["nativeTypes"]
}

datasource db {
  provider = "mysql"
  url      = env("DATABASE_URL")
}

model User {
  id    Int                            @id @default(autoincrement())
  email String                         @unique
  name  String?
  bio   Unsupported("multilinestring") @unique @default(dbgenerated("''"))
}

Prisma Migrate generates the following migration:

CREATE TABLE `User` (
  `id` INTEGER NOT NULL AUTO_INCREMENT,
  `email` VARCHAR(191) NOT NULL,
  `name` VARCHAR(191),
  `multilinestringField` multilinestring NOT NULL DEFAULT '',

  UNIQUE INDEX `User.email_unique`(`email`),
  UNIQUE INDEX `User.bio_unique`(`bio`),
  PRIMARY KEY (`id`)
);

dbgenerated(String)

can also be used with supported native types. Here is an example of how to leverage this to use database-generated UUIDs as default for a UUID column:

model User {
  id   String  @id @db.Uuid @default(dbgenerated("gen_random_uuid()"))
  name String?
}

Prisma Migrate generates the following migration:

CREATE TABLE "User" (
    "id" UUID NOT NULL DEFAULT gen_random_uuid(),
    "name" TEXT,

    PRIMARY KEY ("id")
);

There is a new option

--skip-seeding

for

prisma migrate reset

and

prisma migrate dev

. More info in the release notes for 2.16.0.

<!here> We are looking for a last round of feedback on nativeTypes (supported in Prisma Migrate since 2.15.0). Please share your feedback here or via GitHub. Also see my relevant updates about 2.15.0 and 2.16.0 above ☝️

Is there a way to have prisma show or generate a migration without executing it?

So ah - I think I might have a really dumb question but dropping it in here anyways 😄 I've inherited a Prisma project from another dev and it was on the older prisma betas - no real seed step here, so what I did was to do a

pg_dump / pg_restore

to get a set of data to work with (also, I like developing with a clone of real data). However, when running

prisma migrate

it's always going to reset my db from scratch - is there a good workflow to • Seed my local database with production data • Update my database with some new model (i.e., add a column), without blowing it all up?

Hi, migrate users! There are some potentially breaking changes in the new release, especially if you are using column types that are not the migrate defaults . Also note that the Float <-> Decimal types have been remapped on postgres and mysql. For the whole story, you can read the release notes: https://github.com/prisma/prisma/releases/tag/2.17.0 All these changes are a consequence of the native types preview feature becoming stable: you can express a much richer set of types in your prisma schema now. It's also an important milestone towards the stabilization of migrate. Please feel free to ask any questions and send feedback in this thread or in GitHub issues :)

Hi! I’m currently moving away from self hosted Jenkins (on aws, inside our VPC) to GitHub actions (outside of our VPC). Our database server, RDS, is locked down to only be accessible from our VPC on AWS and it would be nice to keep it this way. Everything works great with GitHub actions except for the step where we need to call Prisma migrate. Does anyone have a similar setup? Can the prisma migrate step happen when the container starts on ECS somehow? I would prefer to do it in the deploy process and preferably without hosting an SSH tunnel on AWS

Hey all! I added the following field to my schema, and am unable to migrate since the existing tables don't have a default value. Is there a way to update existing data with Prisma? New field:

uuid        String             @default(uuid())

Postgres error:

Error: Database error: Null constraint failed: (uuid)