Disks on swarm managers in dev enviroment seem to be filling up often. Often the cause is log files created by Microsoft monitoring agent. Check if its logrotate config is up-to-date: https://learn.microsoft.com/en-us/system-center/scom/manage-monitoring-unix-linux?view=sc-om-2022#example-logrotate-configuration-file
The same issue might be also in prod