This article provides a solution for when Magento cron jobs do not finish executing and persist in a
running status, which prevents other cron jobs from running. This can happen for a number of reasons, such as network issues, application crashes, redeployment issues.
Affected products and versions
Magento Commerce Cloud, all versions
Symptoms of cron jobs that must be reset include:
- Large quantity of jobs appear in the
- Site performance starts to degrade
- Jobs fail to execute on schedule
Running this command without the --job-code option resets all cron jobs, including those currently running, so we recommend using it only in exceptional cases, such as after you have verified that all cron jobs must be reset. Re-deployment runs this command by default to reset cron jobs, so they recover appropriately after the environment is back up. Avoid using this solution when indexers are running.
To resolve this issue, you must reset the cron job(s) using the
cron:unlock command. This command changes the status of the cron job in the database, ending the job forcefully to allow other scheduled jobs to continue.
- Open a terminal and use your SSH keys to connect to the affected environment.
- Get the MySQL database credentials:
shell echo $MAGENTO_CLOUD_RELATIONSHIPS | base64 -d | json_pp
- Connect to the database using
shell mysql -hdatabase.internal -uuser -ppassword main
- Select the
shell use main
- Find all running cron jobs:
shell SELECT * FROM cron_schedule WHERE status = 'running';
- Copy the
job_codeof any job running longer than usual.
- Use the schedule IDs from the previous step to unlock specific cron jobs:
shell ./vendor/bin/ece-tools cron:unlock --job-code=<job_code_1> [... --job-code=<job_code_x>]