Download the PHP package ttree/flowplatformsh without Composer

Platform.sh for Flow Framework made easy

Flow framework utility package to parse Platform.sh variables

Configure your distribution to deploy on Platform.sh

composer require ttree/flowplatformsh
./flow platform:bootstrap --id fajq56c55mc5s --host eu.platform.sh --database MySQL|PostgreSQL

Check and modify the configuration to match your project:

Run a command during Build or Deploy hook

You can run any Flow CLI commands during build or deploy hook. Remember that during Build hook you don't have access to services (database, cache, ...). This the default configuration:

Ttree:
  FlowPlatformSh:
    buildHooks:
      commands:
        'flow:package:rescan': true
    deployHooks:
      commands:
        'flow:cache:flush': true
        'flow:cache:warmup': true
        'flow:doctrine:migrate': true

You can set the value to to disable a command. If the value is an array, the array is passed as command arguments:

Ttree:
  FlowPlatformSh:
   deployHooks:
     commands:
      'ttree.neosplatformsh:platform:createadminaccount':
        username: 'admin'
        password: 'changeme'
      'ttree.neosplatformsh:platform:importsitepackage':
        package: 'Neos.Demo'

Check that your execute and in the respective hook.

You can use a custom FLOW_CONTEXT during the build hook to avoid issue when Flow try to connect to redis cache, like this :

variables:
  env:
    FLOW_CONTEXT: 'Production/PlatformSh'
    FLOW_PATH_TEMPORARY_BASE: '/tmp'
    FLOW_REWRITEURLS: 1

hooks:
  build: |
    set -e
    export FLOW_CONTEXT=${FLOW_CONTEXT}Build
    php flow platform:build
  deploy: |
    set -e
    php flow platform:deploy

Check the command controller in Ttree.NeosPlatformSh.

How to configure

This file extract variables from and create env variables that you can use in your configuration (, , ...). Every line contains the variable name and the path to get the variable content from .

DATABASE_HOST = database.0.host
DATABASE_PORT = database.0.port
DATABASE_NAME = database.0.path
DATABASE_USER = database.0.username
DATABASE_PASSWORD = database.0.password

REDIS_HOST = redis.0.host
REDIS_PORT = redis.0.port

REDIS_ALTERNATIVE_HOST = redis.1.host
REDIS_ALTERNATIVE_PORT = redis.1.port

ELASTICSEARCH_HOST = elasticsearch.0.host
ELASTICSEARCH_PORT = elasticsearch.0.port

Then you can edit your to use the new env variables:

Neos:
  Flow:
    persistence:
      backendOptions:
        driver: pdo_pgsql
        dbname: '%env:DATABASE_NAME%'
        port: '%env:DATABASE_PORT%'
        user: '%env:DATABASE_USER%'
        password: '%env:DATABASE_PASSWORD%'
        host: '%env:DATABASE_HOST%'

Push local data to platform.sh

You can sync a local project data (resources and databases) to platform with the following command:

./flow platform:push --directory Data/Persistent --publish --database --migrate --environment master

You can provide the path to your local with the paramater .

The options run the resources publishing after the rsync command on the remote server.

The options and clone the local database and run migration on the remote server.

The options create a snapshot of the current platform environement before the synchronzation.

The options , default , allow to target specific platform environment.

The options flush all remote caches after the synchronization.

The options preview the commands and execute nothing.

The options non interactive mode.

Warning: Currently we push only files and databases, if you use ElasticSearch you need to rebuild the index manually.

You should see this output:

Local -> platform.sh

    + Create Snapshot
    + Sync directory Data/Persistent
    + Publish resources
    + Clone database
    + Migrate database

Pull data from platform.sh to your local setup

You can sync a local project data (resources and databases) from a platform environment with the following command:

./flow platform:pull --directory Data/Persistent --publish --database --migrate --environment master

You can provide the path to your local with the paramater .

The options run the resources publishing after the rsync command on the remote server.

The options and clone the remote database and run migration on the local server.

The options , default , allow to target specific platform environment.

The options flush all local caches after the synchronization.

The options preview the commands and execute nothing.

The options non interactive mode.

Warning: Currently we pull only files and databases, if you use ElasticSearch you need to rebuild the index manually.

You should see this output:

platform.sh -> Local

    + Sync directory Data/Persistent
    + Publish resources
    + Clone database
    + Migrate database
    + Flush all caches

I use docker locally to host my database, the push command failed

You can use a custom dump command, by editing your , by example for PostgreSQL:

Ttree:
  FlowPlatformSh:
    commands:
      push:
        dump:
          pgsql:
            '*': 'docker exec -e "PGPASSWORD=@PASSWORD@" -t [docker-container-name] pg_dump -c -b -d @DBNAME@ -U @USER@ > Data/Temporary/PlatformShDump-@[email protected]'
      pull:
        restore:
          pgsql:
            '*': 'cat Data/Temporary/PlatformShDump-@[email protected] | docker exec -e "PGPASSWORD=@PASSWORD@" -i [docker-container-name] psql --host=@HOST@ -U@USER@ @DBNAME@'

Replace by the name or the identifier of the PostgreSQL container.

Acknowledgments

Development sponsored by ttree ltd - neos solution provider.

We try our best to craft this package with a lots of love, we are open to sponsoring, support request, ... just contact us.

License

The MIT License (MIT). Please see LICENSE for more information.