tibotiber/hasura-action

GitHub Action wrapping the Hasura CLI

hasura-action's Issues

Avoid printing sensitive info

I saw your comment "secrets can be printed, they are protected by Github Actions", but sometimes they're not. We fetch the Admin secret from AWS Secrets Manager and add it to $GITHUB_ENV.

Unless you have a better approach to this, maybe consider changing the echo statement?

deploy_hasura_configuration:
    name: Hasura Configuration
    needs: deploy_backend
    runs-on: ubuntu-latest
    env:
      HASURA_ENDPOINT: ...
      HASURA_WORKDIR: ...
      # HASURA_ADMIN_SECRET is added to env in the steps below.
    steps:
    - name: Checkout Repo
      uses: actions/checkout@master
    - name: Configure AWS Creds (Development)
      uses: aws-actions/configure-aws-credentials@v1
      with:
        role-to-assume: arn:aws:iam::...
        aws-region: eu-central-1
    - name: Retrieve and set Hasura admin secret
      run: |
        secret_info=$(aws secretsmanager get-secret-value --region ... --secret-id ... --output text)
        read -ra secret_values <<< "$secret_info"
        echo "HASURA_ADMIN_SECRET=${secret_values[3]}" >> $GITHUB_ENV
    - name: Apply database migrations
      uses: tibotiber/hasura-action@master
      with:
        args: migrate apply
    - name: Apply metadata
      uses: tibotiber/hasura-action@master
      with:
        args: metadata apply

Version check and update crashes action

Problem:

Hasura cloud is currently returning 2.15.2 which is not a valid CLI version (yet?), thus the action fails completely for me.

GET https://{...}.hasura.app/v1/version
{"version":"v2.15.2-cloud.1"}

Since the action always looks for an update:

hasura update-cli --version 2.15.2
FATA[0000] apply update: download asset: could not find the release asset

I suggest if this error happens to catch it and try to execute the command anyway.

parse error: Invalid numeric literal at line 1, column 7

Thank you for making this github action. As of ~10 days ago, it has started failing with the following error:
parse error: Invalid numeric literal at line 1, column 7 on running hasura metadata apply

I believe that it's due to taking the latest version of graphql-engine here, and there being a non-backward compatible change in the latest version of hasura graphql (v2.1.0)

Is it possible to use this github action to a fixed version of hasura graphql?

Error when running metadata apply action

I'm getting the following error trying to apply metadata from github actions:

Executing 'hasura metadata apply --endpoint *** --admin-secret ***' from './'
sh: syntax error: unexpected word (expecting ")")

Here's the relevant config:

defaults:
  run:
    shell: bash

jobs:
  update_graphql:
    name: Update Hasura Metadata
    runs-on: ubuntu-latest
    steps:
      - name: Checkout Repo
        uses: actions/checkout@master

      - name: Apply hasura metadata
        uses: tibotiber/hasura-action@master
        with:
          args: metadata apply
        env:
          HASURA_ENDPOINT: ${{ secrets.HASURA_ENDPOINT }}
          HASURA_ADMIN_SECRET: ${{ secrets.HASURA_ADMIN_SECRET }}

Metadata is in the default location ./migrations/metadata.yaml at the root of the project repo.

Running multiple actions

First thanks a ton for making this addon! It has been extremely helpful for testing hasura apps 😄

I'm loading migrations and metadata that are checked into my repo, so in order to get the hasura instance running correctly I have to run: metadata clear, migrate apply, and metadata apply.

I'm wondering if there's a way I can run all three of these using a single step in my yaml file? Right now I have a bit of repetition for these three commands:

- name: Setup Hasura metadata clear
  uses: tibotiber/hasura-action@master
  with:
    args: metadata clear
  env:
    HASURA_ENDPOINT: http://hasura:8080
    HASURA_ADMIN_SECRET: secret
    HASURA_WORKDIR: test/hasura
    HASURA_ENGINE_VERSION: v2.0.10
- name: Setup Hasura migrations
  uses: tibotiber/hasura-action@master
  with:
    args: migrate apply --database-name default
  env:
    HASURA_ENDPOINT: http://hasura:8080
    HASURA_ADMIN_SECRET: secret
    HASURA_WORKDIR: test/hasura
    HASURA_ENGINE_VERSION: v2.0.10
- name: Setup Hasura metadata apply
  uses: tibotiber/hasura-action@master
  with:
    args: metadata apply
  env:
    HASURA_ENDPOINT: http://hasura:8080
    HASURA_ADMIN_SECRET: secret
    HASURA_WORKDIR: test/hasura
    HASURA_ENGINE_VERSION: v2.0.10

Do you have any ideas how I could run all there of these as a single step? If not, would you be open to a PR that allowed this?

Not able to find ACTION_BASE_ENDPOINT env var

I am using Hasura actions with an ACTION_BASE_ENDPOINT. I am trying to create a GitHub workflow using this action. My action code looks like the following:

name: Hasura migration
on:
  push:
    branches:
      - master
jobs:
  hasura_migration:
    name: Hasura migration
    runs-on: ubuntu-latest
    env:
      HASURA_ENDPOINT: ${{ secrets.HASURA_ENDPOINT }}
      HASURA_ADMIN_SECRET: ${{ secrets.HASURA_ADMIN_SECRET }}
      ACTION_BASE_ENDPOINT: ${{ secrets.ACTION_BASE_ENDPOINT }}
      HASURA_WORKDIR: packages/hasura/hasura # replace this by your own path if needed
    steps:
      - name: Checkout Repo
        uses: actions/checkout@master
      - name: Echo env
        run: echo $ACTION_BASE_ENDPOINT $HASURA_ENDPOINT $HASURA_ADMIN_SECRET
      - name: Apply hasura migrations
        uses: tibotiber/hasura-action@master
        with:
          args: migrate apply
      - name: Apply hasura metadata
        uses: tibotiber/hasura-action@master
        with:
          args: metadata apply

When it runs, I get the following error:

HasuraError.internal of type hasuradb.SQLInternalError; response from API: {\"internal\":[{\"definition\":{\"definition\":{\"kind\":\"synchronous\",\"output_type\":\"Result\",\"arguments\":[{\"name\":\"accountData\",\"type\":\"AccountData!\",\"description\":null}],\"headers\":[],\"handler\":\"{{ACTION_BASE_ENDPOINT}}/link-account\",\"type\":\"mutation\",\"forward_client_headers\":true},\"name\":\"linkAccount\",\"comment\":null},\"reason\":\"in action \\\"linkAccount\\\"; Value for environment variables not found: ACTION_BASE_ENDPOINT\",\"type\":\"action\"},

I don't understand why the ACTION_BASE_ENDPOINT var isn't being set. Does it need to be passed through to the script within the workflow action itself?

Error Missing --database-name flag for V3

Hi, I'm trying to setup github action with this action but I'm having this issue saying --database-name flag is required. Someone know what's the potential issue?

time="2022-05-25T16:26:56Z" level=fatal msg="--database-name flag is required"

--database-name is now a required field.
https://hasura.io/docs/latest/graphql/core/migrations/upgrade-v3/#changes-needed-in-existing-workflows

Segfault using this action

{"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/1.0/graphql/manual/guides/telemetry.html\n","time":"2020-07-28T19:05:02Z"}
{"level":"info","msg":"Checking for update... ","time":"2020-07-28T19:05:02Z"}
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x758106]

goroutine 1 [running]:
github.com/Masterminds/semver.(*Version).Major(...)
	/go/pkg/mod/github.com/!masterminds/[email protected]/version.go:131
github.com/Masterminds/semver.(*Version).Compare(0x0, 0xc0000beeb0, 0x0)
	/go/pkg/mod/github.com/!masterminds/[email protected]/version.go:268 +0x26
github.com/Masterminds/semver.(*Version).GreaterThan(...)
	/go/pkg/mod/github.com/!masterminds/[email protected]/version.go:250
github.com/hasura/graphql-engine/cli/update.HasUpdate(0xc0000beeb0, 0xc0003ef770, 0x29, 0xfb1300, 0x0, 0xc0002b1a00, 0x0, 0x0, 0x0)
	/root/graphql-engine/cli/update/update.go:98 +0x12d
github.com/hasura/graphql-engine/cli/commands.(*updateOptions).run(0xc000561c70, 0x1, 0xc00031dc01, 0xc6f497)
	/root/graphql-engine/cli/commands/update-cli.go:75 +0x5fc
github.com/hasura/graphql-engine/cli/commands.glob..func1(0xc000169b80, 0xc0000b5ac0, 0x0, 0x4)
	/root/graphql-engine/cli/commands/root.go:42 +0xc3
github.com/hasura/graphql-engine/cli/commands.NewMetadataCmd.func1(0xc000169b80, 0xc0000b5ac0, 0x0, 0x4, 0x0, 0x0)
	/root/graphql-engine/cli/commands/metadata.go:20 +0x79
github.com/spf13/cobra.(*Command).execute(0xc000169b80, 0xc0000b5a80, 0x4, 0x4, 0xc000169b80, 0xc0000b5a80)
	/go/pkg/mod/github.com/spf13/[email protected]/command.go:805 +0x56b
github.com/spf13/cobra.(*Command).ExecuteC(0x19da780, 0xc00016dbc0, 0xc0000b6000, 0x7)
	/go/pkg/mod/github.com/spf13/[email protected]/command.go:914 +0x2fb
github.com/hasura/graphql-engine/cli/commands.Execute(0xc000084058, 0x0)
	/root/graphql-engine/cli/commands/root.go:119 +0x13a
main.main()
	/root/graphql-engine/cli/cmd/hasura/hasura.go:11 +0x22

keeping CLI in sync with server

I think it is possible to detect the version of the user's target hasura endpoint https://hasura.io/docs/latest/graphql/core/api-reference/version.html#endpoint then make the action pull the correct version of the cli. This would allow users to not worry about changing the HASURA_ENGINE_VERSION configuration if they want to update their server.

Errors applying metadata

I've got this running fine for applying migrations, but when applying metadata I get this error: level=fatal msg="failed to apply metadata: cannot apply metadata on the database: cannot build actions from metadata: error in converting sdl to metadata: fork/exec /github/home/.hasura/plugins/bin/hasura-cli_ext: no such file or directory:".

I think it's because the Hasura CLI is failing to install the plugin needed for parsing metadata, but I'm not sure. Is this something you've encountered?

Actually the logs seem to suggest the plugin is successfully installed, but the Hasura CLI can't find it.

Action recently stopped installing

  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
100   139  100   139    0     0   1561      0 --:--:-- --:--:-- --:--:--  1561
--> Getting latest version...

100  3482  100  3482    0     0  19131      0 --:--:-- --:--:-- --:--:-- 19131
--> 
--> Failed while attempting to install hasura graphql-engine cli. Please manually install:
--> 
--> 2. Open your web browser and go to https://github.com/hasura/graphql-engine/releases
--> 2. Download the cli from latest release for your platform. Name it 'hasura'.
--> 3. chmod +x ./hasura
--> 4. mv ./hasura /usr/local/bin
--> 
FAILED> exiting... 
The command '/bin/sh -c curl -L https://github.com/hasura/graphql-engine/raw/stable/cli/get.sh | bash' returned a non-zero code: 1

##[error]Docker build failed with exit code 1

Avoid downloading the action when skipped

Hey!

As you can see on this image, both operations, Apply migrations and Apply metadata, were skipped due to having no changes, yet the action proceeded to download (and take 23s) for an action that should've taken 2s or 3s.

How can I avoid this?

Thank you!

Applying metadata segfaults

We see segmentation fault when applying metadata from a github action:

{"level":"info","msg":"Applying metadata...","time":"2022-01-10T17:45:01Z"}
21
time="2022-01-10T17:45:02Z" level=fatal msg="error applying metadata \ncannot build actions from project: error parsing metadata \nobject: actions\nfile: actions.yaml\nerror: error in converting sdl to metadata: signal: segmentation fault (core dumped): /github/home/.hasura/cli-ext/v2.0.9/cli-ext-730925829/cli-ext: /usr/lib/libstdc++.so.6: no version information available (required by /github/home/.hasura/cli-ext/v2.0.9/cli-ext-730925829/cli-ext

Digging into the build of the action container I see

2022-01-10T17:44:44.0128890Z Step 9/14 : RUN apk add --no-cache curl bash libstdc++ jq
2022-01-10T17:44:44.0305814Z  ---> Running in f22c095441c3
2022-01-10T17:44:44.3144479Z fetch https://dl-cdn.alpinelinux.org/alpine/v3.15/main/x86_64/APKINDEX.tar.gz
2022-01-10T17:44:44.4449166Z fetch https://dl-cdn.alpinelinux.org/alpine/v3.15/community/x86_64/APKINDEX.tar.gz
2022-01-10T17:44:44.6575350Z (1/12) Installing ncurses-terminfo-base (6.3_p20211120-r0)
2022-01-10T17:44:44.6727502Z (2/12) Installing ncurses-libs (6.3_p20211120-r0)
2022-01-10T17:44:44.6910963Z (3/12) Installing readline (8.1.1-r0)
2022-01-10T17:44:44.7070337Z (4/12) Installing bash (5.1.8-r0)
2022-01-10T17:44:44.7328750Z Executing bash-5.1.8-r0.post-install
2022-01-10T17:44:44.7353334Z (5/12) Installing ca-certificates (20191127-r7)
2022-01-10T17:44:44.7633146Z (6/12) Installing brotli-libs (1.0.9-r5)
2022-01-10T17:44:44.7821824Z (7/12) Installing nghttp2-libs (1.46.0-r0)
2022-01-10T17:44:44.7958302Z (8/12) Installing libcurl (7.80.0-r0)
2022-01-10T17:44:44.8135866Z (9/12) Installing curl (7.80.0-r0)
2022-01-10T17:44:44.8293533Z (10/12) Installing oniguruma (6.9.7.1-r0)
2022-01-10T17:44:44.8505042Z (11/12) Installing jq (1.6-r1)
2022-01-10T17:44:44.8725818Z (12/12) Installing libstdc++ (10.3.1_git20211027-r0)
2022-01-10T17:44:44.9007027Z Executing busybox-1.34.1-r3.trigger
2022-01-10T17:44:44.9043636Z Executing ca-certificates-20191127-r7.trigger
2022-01-10T17:44:44.9411775Z Executing glibc-bin-2.34-r0.trigger

wheres our last successful run had

2022-01-10T07:32:01.0115679Z Step 9/14 : RUN apk add --no-cache curl bash libstdc++ jq
2022-01-10T07:32:01.0374367Z  ---> Running in 83ba1d8d1007
2022-01-10T07:32:01.3500892Z fetch https://dl-cdn.alpinelinux.org/alpine/v3.15/main/x86_64/APKINDEX.tar.gz
2022-01-10T07:32:01.4700518Z fetch https://dl-cdn.alpinelinux.org/alpine/v3.15/community/x86_64/APKINDEX.tar.gz
022-01-10T07:32:01.7084359Z (1/12) Installing ncurses-terminfo-base (6.3_p20211120-r0)
2022-01-10T07:32:01.7207163Z (2/12) Installing ncurses-libs (6.3_p20211120-r0)
2022-01-10T07:32:01.7355017Z (3/12) Installing readline (8.1.1-r0)
2022-01-10T07:32:01.7554953Z (4/12) Installing bash (5.1.8-r0)
2022-01-10T07:32:01.7791060Z Executing bash-5.1.8-r0.post-install
2022-01-10T07:32:01.7811632Z (5/12) Installing ca-certificates (20191127-r7)
2022-01-10T07:32:01.8028805Z (6/12) Installing brotli-libs (1.0.9-r5)
2022-01-10T07:32:01.8201841Z (7/12) Installing nghttp2-libs (1.46.0-r0)
2022-01-10T07:32:01.8374849Z (8/12) Installing libcurl (7.80.0-r0)
2022-01-10T07:32:01.8523015Z (9/12) Installing curl (7.80.0-r0)
2022-01-10T07:32:01.8646382Z (10/12) Installing oniguruma (6.9.7.1-r0)
2022-01-10T07:32:01.8799539Z (11/12) Installing jq (1.6-r1)
2022-01-10T07:32:01.9002994Z (12/12) Installing libstdc++ (10.3.1_git20211027-r0)
2022-01-10T07:32:01.9819847Z Executing busybox-1.34.1-r3.trigger
2022-01-10T07:32:01.9859895Z Executing ca-certificates-20191127-r7.trigger
2022-01-10T07:32:02.0250041Z Executing glibc-bin-2.33-r0.trigger
2022-01-10T07:32:02.0266519Z �/usr/glibc-compat/sbin/ldconfig: /usr/glibc-compat/lib/ld-linux-x86-64.so.2 is not a symbolic link

Full failed log here

So I think the github action environment got updated and broke things.