We noticed this issue after upgrading from 5.2.8 to 5.2.9 on a number of 60Ds. While working with Fortinet support, he said it's a known issue with the 60Ds and 90Ds with how the IPS engine in 5.2.9 works with the hardware.
The bundled IPS engine version is 3.00170, and when working with support they uploaded a new IPS def to bring the engine to version 3.00172. That corrected the issue for us, the ipsengine process hasn't crashed for going on 24 hours now (had been crashing every minute). If you have support, open a ticket to get the new engine.
According to support the engine is only updated manually by uploading the file (or when you upgrade code as it's bundled in). I'm not sure how granular the def updates are, but this engine is specific to the 60D and 90D models. Our firewalls check for updates at least every two hours and over 24 hours didn't auto pull the new engine; manually telling it to update didn't update the engine either after waiting about an hour.
I haven't found the engine listed on the support site.
We run pretty much full UTM on all connections. In our case, the 60Ds we saw the crashing issue at ~1000 sessions, whereas the 60Ds not processing as many sessions either didn't have the crashing issue or it was much less frequent. You can check for the crashing under Log & Report > System or in an Analyzer if you have one.
The Fortinet Security Fabric brings together the concepts of convergence and consolidation to provide comprehensive cybersecurity protection for all users, devices, and applications and across all network edges.