If a backup/snapshot activity appears to be stuck, there could be multiple causes. Generally, they can be similar to stuck deployment activities. For instance, the automation system might be:
- Waiting for activities to finish first, e.g. a long-running cron job.
- Hitting a bug such as having difficulties to unfreeze a mounted disk after taking a snapshot.
Long-running jobs on the environment will mark the backup activity as pending with a message such as the one below.
Waiting for other activities to complete
The job in question can be found as follows.
1. SSH into the environment:
$ platform ssh
2. The long-running cron jobs on the environment can be found with the command:
$ ps afx
3. Once the long-running process has been identified (<PID>), it can be killed with the command:
$ kill <PID>
Unfortunately, it is not yet possible to cancel a backup activity. Should the above not solve your issue, please open a support ticket with our 24/7 support team via your project console.
Alternatively to regular backups, you may want to try to utilize Live Backups, however those might result in inconsistent data as data integrity can not be guaranteed with this method.
You can read up more about the topic of backups on Platform.sh in general as well as recommended steps to solve stuck deployments, which might give further clues, following the links below:
- https://docs.platform.sh/environments/backup.html
- https://docs.platform.sh/development/troubleshoot.html#stuck-build-or-deployment
- https://docs.platform.sh/development/troubleshoot.html#cron-jobs