How to merge two RAID arrays?

Question

RAID

Hello, I have a DL380gen9 server with two drive cages. There are 12x2 8TB drives installed in the baskets = 192TB, the baskets are connected via the Smart Array p841 controller. Inside the server, there are 7 more of the same disks, 8 each. But they are connected to another Smart Array p840 raid controller.
The question is how to combine them into one raid array? I did it using Windows functions, created a composite volume from two RAID0 arrays, how can I do it differently? What problems might arise? If you collect RAID5-6 arrays and create a spanned volume in the same way, is it possible to hot-swap in the future if one of the disks fails?
All this for the maximum possible file storage space, speed is not so critical.

Answer 1

If disks are connected to different controllers, then you cannot combine them into one RAID at the controller level. By the way, it is better to have several raids of a smaller size than one for all the available space.

Answer 2

spanned volume from two RAID0 arrays... What problems can arise?

In my opinion, it is important to remember here that RAID 0 of N disks is N times less reliable than a single disk. If you lose one disk in such a raid, you will lose everything.
Usually, using REID 0 in some more or less serious work scenario makes sense only when speed takes precedence over everything else.

Answer 3

There is not enough data to solve the problem. But the first thing that comes to mind is to connect all the disks to one controller. There are most likely 4 connectors on the P841, if they are not all occupied, then we take one of them and connect it inside the server, to the backplain. And P840 is simply thrown out of the server.
That is, we wrap the cable from the external connector carefully through the plug inward. Not exactly Feng Shui, but a working solution. You can connect all disks to one controller and build one large array. Just imagine that the front of the server with the disks and the backplane is essentially the same disk shelf.
Here, however, you need to look at the points of failure both in the server and in the shelves.
Ideally, I would already take HBA for such a number of disks and connect all all disks to it, and then a software raid, depending on the OS and task.