Sheduled Scan taking hours

Hi, we are investigating CPU spikes on our multisite WP. I could recently trace the spikes to the WordFence scan which runs periodically every 2-4 days. The problem is that these spikes take several hours. Ranging from 6 to 24 hours based on the logs. In that time one core is also (obviously) blocked for this script. I’m totally aware that this should take some time, as the scan has to check many things. However we are running an installation with about 4000 pages and this number is expected to increase far beyond that. Using the trace command I was able to identify alot of blog switches from the WF scan process instance. This would mean that with increasing blog count I have to expect the scan time to increase to way more than a day (as this has already been reached once).

So: Is this the expected time frame for a scan with “just” 4000 subblogs? And can I do something about it? Like, I’m thinking about buying pro the reshedule the scan to the weekend, to ahve a little more control. However if 4000 pages can take up to 24 hours, simply getting to 8000 would mean it already could take 2 days. So it’s not that long until just one scan is longer than the weekend itself.

PS: We are not experiencing any downtimes or performance issues (for now), but I’m concerned with basically having one core just reserved for continuous scans over multiple days, when the site reaches medium traffic.