Dear community,
We are implementing new NVMe disks in our cluster and currently discussing about the best redundancy methods. As these disks are not meant to be managed by a hardware controller, we have multiple disks (in fact 4), that can only be mounted as separate disks into the VM of a FortiSIEM worker.
We are thinking about mounting two or four disks on the Rocky Linux of the worker, then using a software solution to use these disks as a raid and then inserting that into the configuration of FortiSIEM (as Clickhouse Hot Node).
Has anyone experience in doing so? If yes, how would that behave on a FSM update?
The first alternative we came up with, would be to setting up multiple workers, one per disk, and getting the redundancy in that way, but losing some disk space and some performance.
Appreciate any thoughts!
Best,
Christian
FCP & FCSS Security Operations | Fortinet Advanced Partner
