Following tools are provided by Pimcore to support deployment processes.
All Pimcore configurations are saved as YAML or PHP files on the file system. As a result they can be included into version control systems and by utilizing the multi environment feature different configuration files for different deployment stages can be defined.
Pimcore Class Definitions
As with Pimcore configurations also Pimcore class definitions are saved as PHP configuration files and therefore can be added to version control systems and be deployed to different deployment stages.
The PHP configuration files and PHP classes will be written to the
var/classes directory by default.
To disallow modification and turn a class to be read-only, you can create a copy
Regarding the class modification, there is also an optional env variable
PIMCORE_CLASS_DEFINITION_WRITABLE that can be considered and set.
0To disallow completely write access, including the creation of new classes.
1To allow the modification, including the classes in
config/pimcore/classesthat normally are read-only.
not setclasses in
config/pimcore/classesare read-only, but new classes are allowed and will be created in
With the env variable
PIMCORE_CLASS_DEFINITION_DIRECTORY you can specify the directory to search for your class definitions
if you do not want pimcore to search in
Note: Changes on Pimcore class definitions not only have influence to configuration files but also on the database. If deploying changes between different deployment stages also database changes need to be deployed. This can be done with the
After every code update you should use the
pimcore:deployment:classes-rebuild command to push changes to the database.
To create new classes from your configuration files in the database you can use the
./bin/console pimcore:deployment:classes-rebuild --create-classes
As an alternative also class export to json-files and the class import commands can be used.
./bin/console pimcore:definition:import:objectbrick /brick_jsonfile_path.json
./bin/console pimcore:definition:import:fieldcollection /collection_jsonfile_path.json
./bin/console pimcore:definition:import:class /class_jsonfile_path.json
The Pimcore Console provides several useful tasks for deployment. These tasks can be integrated into custom deployment workflows and tools. One example for them would be the Pimcore class definitions as described above.
To get a list of all available commands use
Potentially useful commands:
|pimcore:mysql-tools||Optimize and warm up mysql database|
|pimcore:search-backend-reindex||Re-indexes the backend search of Pimcore (only available if you have installed the simpleBackendSearchBundle)|
|pimcore:cache:clear||Clear Pimcore core caches|
|cache:clear||Clear Symfony caches|
|pimcore:cache:warming||Warm up caches|
|pimcore:classificationstore:delete-store||Delete Classification Store|
|pimcore:definition:import:class||Import Class definition from a JSON export|
|pimcore:definition:import:customlayout||Import Customlayout definition from a JSON export|
|pimcore:definition:import:fieldcollection||Import FieldCollection definition from a JSON export|
|pimcore:definition:import:objectbrick||Import ObjectBrick definition from a JSON export|
|pimcore:deployment:classes-rebuild||Rebuilds classes and db structure based on updated |
|pimcore:thumbnails:image||Generate image thumbnails, useful to pre-generate thumbnails in the background. Use |
|pimcore:thumbnails:optimize-images||Optimize file size of all images in |
|pimcore:thumbnails:video||Generate video thumbnails, useful to pre-generate thumbnails in the background. Use |
Find more about the Pimcore Console on the dedicated page.
The content migration between environments is not provided by Pimcore and it's not recommended at all.
The content should be created by editors in the production environment and visibility on the frontend can be managed by built-in features like publishing / unpublishing / versioning / scheduling / preview the effect in editmode.
Therefore, editors shouldn't work on different stages.
Of course, the content migration is possible but this is always a very individual task depending on data model, environments and use cases.