In the machine where I keep the backups for when a datacenter burns down I have an older Areca RAID controller (ARC-1231ML).
For some reason, it always runs a check on all volumes directly after reboot, My one suspicion is that this is triggered after it updates its RTC using NTP, which causes a long time skip, so the scheduled time is immediately reached, the other is that a check is always run if no BBU is installed (it died a while ago).
Is there a way to avoid the immediate check?
Likely, the dead BBU is the problem. Without it, the RAID controller doesn't know the state of the array after a reboot and needs to verify it. You should replace the BBU or the entire controller.