There is a Synology RS2818RP+, how best to implement fault tolerance?

Question

Storage Area Network

Good afternoon, there is RS2818RP + 16 disks of 4TB in raid 10. 20 TB of information has accumulated. The moons are connected via iscsi to virtual machines. Because it takes a very long time to roll out backups, I am considering purchasing another RS2818RP + for organizing HA. For communication with servers via iscsi 10GB over twisted pair, separate switches (san).

C HA worked on Tekus, the experience was sad, because. when at one point there was a failure, the tekus switched to passive storage, and when it rebooted the main one, everything became ok, but the tekus began to synchronize the cluster with each other at a speed of 1 MB / s when accessing the storage))))))))))) ))))))))))))))))) And they had 100 TB of information ...

1. The question is, does it work normally on synology? I am primarily interested in iscsi .
2. How quickly will the "copying" to the passive storage happen? (on the main already 20TB).
3. In another RS2818RP+, I want to increase the capacity, can I change disks one by one for a larger volume, each time restoring the raid and eventually get a larger volume?
4. Do you think that when using HA, do you need a 10 raid or can you get by with a ssd cache + raid 6+ disk replacement? Write/read speed is important.
5. Or is it better to take a two-headed shelf instead of HA?
6. Or in general synology in the furnace, at the previous work was lenovo with FC, thecus's in HA, synology. Complaints were only on the current, slow read / write speed and HA was on the first failure.

Answer 1

4. Do you think that when using HA, do you need a 10 raid or can you get by with a ssd cache + raid 6+ disk replacement? Write/read speed is important.
in the sense that it is enough to maintain the current performance.