Hasura with Remix on Fly

Install Hasura CLI if you haven't already: https://hasura.io/docs/latest/graphql/core/hasura-cli/install-hasura-cli/

To run Hasura locally, just run docker-compose up. To access the Hasura console, run hasura console from the hasura directory.

Initial setup might include running this:

hasura metadata apply
hasura migrate apply
hasura metadata reload
hasura seed apply

Linux support

This works "out-of-the-box" on Arch Linux, probably on every other distribution too.

Only thing you have to change is docker-compose.yml.

You'll have to change 3 lines manually. Terrifying, I know. To make it easier for you, I've marked the lines with a comment.

Rest of this document is about getting it running on Fly and your Remix application talking to it.

Setting it up

Setup the Remix application before proceeding with this. And don't use hunter1 as your admin secret!

$ fly create remix-prog-stack-hasura
$ fly postgres create --name remix-prog-stack-hasura-db
$ fly postgres attach --postgres-app remix-prog-stack-hasura-db --app remix-prog-stack-hasura

$ fly secrets set HASURA_GRAPHQL_DATABASE_URL="[DATABASE_URL_FROM_ATTACH]" --app remix-prog-stack-hasura

$ fly secrets set HASURA_GRAPHQL_ADMIN_SECRET="hunter1" --app remix-prog-stack-hasura
$ fly secrets set HASURA_GRAPHQL_AUTH_HOOK="https://remix-prog-stack.fly.dev/hasura-auth" --app remix-prog-stack-hasura

# Before running this, go and change the memory limit to 512. Otherwise your deploy will fail.
# https://fly.io/apps/remix-prog-stack-hasura/scale
$ fly deploy

Then configure the application so it so it can talk to Hasura.

fly secrets set HASURA_URL="https://remix-prog-stack-hasura.fly.dev/v1/graphql" --app remix-prog-stack

# This should also work, but it didn't for me.
# fly secrets set HASURA_URL="http://remix-prog-stack-hasura.internal/v1/graphql" --app remix-prog-stack

fly secrets set HASURA_ADMIN_SECRET="hunter1" --app remix-prog-stack

Note: If you do this from the hasura directory, you get a similar warning as shown earlier. It's probably fine to say yes. Alternatively, navigate to the application directory to set the secrets.

Seeding the database

hasura seed apply --endpoint https://remix-prog-stack-hasura.fly.dev

To seed the local database, just leave the endpoint parameter out.

More info on https://hasura.io/docs/latest/graphql/core/hasura-cli/hasura_seed/

Migrations

As long as you do your edits through hasura console, migrations should sort themselves out on fly deploy, assuming the process doesn't run out of memory... If you don't need the extra juice and wish to save some money, edit the Dockerfile so it doesn't use a cli-migrations version of Hasura, and run the migrations manually.

To apply the migrations manually:

#!/bin/sh
# migrate.sh

echo "Applying migrations on https://remix-prog-stack-hasura.fly.dev"

hasura metadata apply --endpoint https://remix-prog-stack-hasura.fly.dev
hasura migrate apply --all-databases --endpoint https://remix-prog-stack-hasura.fly.dev
hasura metadata reload --endpoint https://remix-prog-stack-hasura.fly.dev

VM Requirements

Hasura can work with the smallest available preset, shared cpu and 256MB of RAM. However, migrations will not.

If you see something like this in a failed deployment, increase the memory.

2022-06-18T21:55:12Z [info]{"timestamp":"2022-06-18T21:55:12.000+0000","level":"info","type":"startup","detail":{"kind":"migrations-apply","info":"applying metadata from /hasura-metadata"}}

2022-06-18T21:55:13Z [info][ 3.657854] **Out of memory: Killed process 523 (graphql-engine) total-vm:1074264644kB, anon-rss:134036kB, file-rss:4kB, shmem-rss:0kB, UID:0 pgtables:404kB oom_score_adj:0

2022-06-18T21:55:13Z [info]{"level":"info","msg":"Help us improve Hasura! The cli collects anonymized usage stats which\nallow us to keep improving Hasura at warp speed. To opt-out or read more,\nvisit https://hasura.io/docs/latest/graphql/core/guides/telemetry.html\n","time":"2022-06-18T21:55:13Z"}

2022-06-18T21:55:13Z [info]{"level":"error","msg":"connecting to graphql-engine server failed","time":"2022-06-18T21:55:13Z"}

Don't be fooled by what's under it:

2022-06-18T21:55:13Z [info]{"level":"info","msg":"possible reasons:","time":"2022-06-18T21:55:13Z"}

2022-06-18T21:55:13Z [info]{"level":"info","msg":"1) Provided root endpoint of graphql-engine server is wrong. Verify endpoint key in config.yaml or/and value of --endpoint flag","time":"2022-06-18T21:55:13Z"}

2022-06-18T21:55:13Z [info]{"level":"info","msg":"3) Server might be unhealthy and is not running/accepting API requests","time":"2022-06-18T21:55:13Z"}

t should be: https://hasura-cloud-app.io","time":"2022-06-18T21:55:13Z"} 2022-06-18T21:55:13Z [info]{"level":"info","msg":"3) Server might be unhealthy and is not running/accepting API requests","time":"2022-06-18T21:55:13Z"}

2022-06-18T21:55:13Z [info]{"level":"info","msg":"4) Admin secret is not correct/set","time":"2022-06-18T21:55:13Z"}

2> 022-06-18T21:55:13Z [info]{"level":"info","msg":"","time":"2022-06-18T21:55:13Z"} 2022-06-18T21:55:13Z [info]time="2022-06-18T21:55:13Z" level=fatal msg="making http request failed: Get "http://localhost:9691/v1/version\": read tcp 127.0.0.1:45740->127.0.0.1:9691: read: connection reset by peer"

2022-06-18T21:55:13Z [info]Main child exited normally with code: 1

2022-06-18T21:55:13Z [info]Starting clean up.

--> v19 failed - Failed due to unhealthy allocations - not rolling back to stable job version 19 as current job has same specification and deploying as v20