On this page
There are times where you might want to trigger a redeployment of your application, such as to add custom TLS certificates. A redeploy reuses your built app and services.
To trigger a redeploy, follow these steps:
The redeploy takes place after any scheduled activities (either Running or Pending).
Despite the name, redeployment doesn’t rerun the
deploy hook, only the
deploy hooks are tied to individual commits in code.
They’re reused until another commit is pushed to the environment.
See more about hooks and their reuse.
To rerun the
deploy hooks, manually trigger a build.
To increase performance and keep applications the same across environments, Platform.sh reuses built applications if its code and build time configuration (variables and such) remain the same.
There may be times where you want to force your application to be built again without changing its code, for example to test an issue in a build hook or when external dependencies change. To force a rebuild without changing the code, use an environment variable.
Assuming you want to do this for your
first create a
REBUILD_DATE environment variable:
platform variable:create --environment main --level environment --prefix env --name REBUILD_DATE --value "$(date)" --visible-build true
This triggers a build right away to propagate the variable. To force a rebuild at any time, update the variable with a new value:
platform variable:update --environment main --value "$(date)" "env:REBUILD_DATE"
This forces your application to be built even if no code has changed.
You may find that you need to clear the build cache, such as when it’s grown too big or, in rare circumstances, when it’s corrupted. It may get corrupted when code is downloaded from a third-party language service like Packagist or npm while that service is experiencing issues.
To clear the build cache, run the following command:
The next build for each environment is likely to take longer as the cache rebuilds.
In most cases, issues accessing a project are caused by missing permissions for a given user. For more information see how to manage user permissions.
If you are using the CLI, make sure that you are authenticated.
If you are using SSH, see how to troubleshoot SSH access.
If you see these errors when accessing your application, it indicates your application is crashing or unavailable.
Typical causes and potential solutions include:
- Your app is listening at the wrong place.
.platform.app.yamlconfiguration has an error and a process isn’t starting or requests can’t be forwarded to it correctly.
- Check your
web.commands.startentry or your
- Check your
- The amount of traffic coming to your site exceeds the processing power of your application.
- Certain code paths in your application are too slow and timing out.
- Check your code is running smoothly.
- Consider adding an observability solution to get a better view of your application.
- A PHP process is crashing because of a segmentation fault.
- A PHP process is killed by the kernel out-of-memory killer.
If you can’t access some part of your project, whether it’s the live site, development environment, or Console, check the Platform.sh status page. There you can see planned maintenance and subscribe to updates for any potential outages.
If the status is operational, contact support.
When you’ve added a command line tool (such as Drush), you might encounter an error like the following:
-bash: drush: command not found
If you see this, add the command you want to run to your path
.environment file script.
As a Linux or Unix-like operating system user (MacOS included),
to be able to run your scripts directly and quickly get past this error
you may need to run the
chmod +x YOUR_SCRIPT_FILE_NAME command.
However, regardless of which operating system you’re using, it’s best if you don’t rely on scripts having execute permissions. Instead, call the app/shell/runtime directly passing your script file to that executable.
If you push code to Platform.sh without the full Git history, sometimes commits are missing. This can happen if you’re pushing code from an external CI/CD pipeline, such as a GitHub action. Such pipelines often do only shallow clones by default.
In such cases, your build might fail with an internal error.
Or you might see an error like
unexpected disconnect while reading sideband packet.
To avoid the error, make sure you do a full clone of the repository before pushing code.
For example, for the Checkout GitHub action,
fetch-depth: 0 to clone the full history.
For GitLab, set clones to have a limit of
0 either in repository settings
or using the
When trying to upload a large JSON file to your API, you might see a 400 response code (
Platform.sh enforces a 10 MB limit on files with the
To send large files, use the
multipart/form-data header instead:
curl -XPOST 'https://example.com/graphql' --header 'Content-Type: multipart/form-data' --form file=large_file.json
For MySQL specific errors, see how to troubleshoot MySQL.
If you try to use a user to create a database, you get an error saying
permission denied to create database.
The database is created for you
and can be found in the
path key of the
PLATFORM_RELATIONSHIPS environment variable.
If you attempt to write to disk outside a
build hook, you may encounter a
read-only file system error.
Except where you define it, the file system is all read-only, with code changes necessary through git.
This gives you benefits like repeatable deployments, consistent backups, and traceability.
To generate anything you need later, write to disk during a
Or declare mounts,
which are writable even during and after deploy.
They can be used for your data: file uploads, logs, and temporary files.
You might see the following message when attempting to run
There isn't enough free space to complete the push
This usually indicates that large files are present in the repository (where they shouldn’t be).
Make sure that the paths for files like media files, dependencies, and databases are set to be ignored in your
If large files are already in the repository, the open-source tool bfg-repo-cleaner can help in cleaning up the repository by purging older commits, removing unnecessary files, and more.
If none of these suggestions work, open a support ticket.
If you see a build or deployment running longer than expected, it may be one of the following cases:
- The build is blocked by a process in your
- The deployment is blocked by a long-running process in your
- The deployment is blocked by a long-running cron job in the environment.
- The deployment is blocked by a long-running cron job in the parent environment.
To determine if your environment is being stuck in the build or the deployment, check your activity log.
If the activity has the result
success, the build has completed successfully and the system is trying to deploy.
If the result is still
running, the build is stuck.
In most regions, stuck builds terminate after one hour.
When a deployment is blocked, you should try the following:
- Connect to your environment using SSH.
- Find any long-running cron jobs or deploy hooks on the environment by running
- Kill any long-running processes with
kill PID. Replace
PIDwith the process ID shown by
activate process is stuck, try the above on the parent environment.
Note that, for PHP apps, you can restart processes that get stuck during a build or deployment from your app container.
Builds can take long time or fail. Most of the time, it’s related to an application issue. Here are a few tips that can help you find the exact cause.
Invisible errors during the build and deploy phase can cause increased wait times, failed builds, and other problems. Investigate each log and fix any errors you find.
deploy hooks can cause long build times.
If they run into issues, they can cause the build to fail or hang indefinitely.
build hooks can be tested in your local environment.
deploy hooks can be tested either locally
or by logging into the application over SSH and running them there.
Be careful not to test the scripts on production environments.
You can also test your hooks with these Linux commands to help debug issues:
time YOUR_HOOK_COMMAND # Print execution time
strace -T YOUR_HOOK_COMMAND # Print a system call report
Containers can’t be shutdown while long-running cron jobs and scheduled tasks are active. That means long-running cron jobs block a container from being shut down to make way for a new deploy.
Make sure your custom cron jobs run quickly and properly. Cron jobs may invoke other services in unexpected ways, which can increase execution time.
A common source of performance issues is a misconfigured cache. The most common issue isn’t allowing the right cookies as part of the router cache.
Some cookies, such as session cookies, need to be allowed. Others, such as marketing and analytics cookies, usually shouldn’t be part of the cache key.
Because the router cache follows cache headers from your app,
your app needs to send the correct
For static assets, set cache headers using the
expires key in your app configuration.
For language-specific troubleshooting for your apps: