Yes, you will want to align to the stripe width of your RAID array just like what is done for SSD's erase block size.
Like this will write in blocks of 512K however your RAID may be set larger then that. If you are not sure I'd go with 512KB's.
mkfs.ext4 -b 4096 -E stride=128,stripe-width=128 /dev/XXXX
(-b 4096 == 4KB blocks) * 128 == 512KB's written at a time.
Both fdisk and gdisk will align your partitions to 1MB by default which will work very well I am sure.
]]>What filesystem would suit best for my use? ext4?
Second: block size; in NTFS one can set the block size, would it make sense to set it very large (because of the large files)? Would it make sense to set it e.g. to 4 MB? Can I do this in ext3/4?
]]>