Heads up folks at VMware,
there are still (many) situations where the vPostgres DB will not clean up deleted rows or tables are flodded with event data causing vpxd to stop working once the vpostgres db disk filled up.
The kb's available are not updated for vCSA 6.0 and are no longer valid since vCSA 6.0 will have different locations for the vPostgres config file (/storage/db) and the actual data files (/storage/seat) and I was unable to find out where in the init scripts the data-location is configured. In addition the postgres / vpostgres user roles and permissions seem to have changed from earlier vCSA releases and e.g. the vacuum command will not accept the password stored in the vpostgres config file.
Thus all available kbs for handling manual cleanup are no longer correct and need to be updated.
I just spend 5 hours fixing the mess caused by one lousy component.
relevant (but outdated) kbs:
VMware KB: Adding additional temporary storage space to vCenter Server Appliance for VMware Postgres maintenance
VMware KB: Increase the disk space in vCenter Server Appliance