Hi! I've actually ran into issue this week related to this topic. I haven't found a solution yet and the testing is still ongoing. I work in an environment where our IIS servers have to be configured according to what is put forth by DISA STIGS for IIS, specifically the part about having to use memory values other than 0 for both private and memory limits. The problem we are running into is something I caught during a migration going from on-prem to DISA cloud. The on-prem IIS web server is built in a VMWARE hypervisor. The IIS web server on the cloud is built in a KVM hypervisor. Both IIS web servers are hosting the same application and using the same IIS configuration settings, and same RAM and app pool memory limits. However, the difference is that the IIS web server that is using the KVM hypervisor is posting every minute for all app pools event ID 5077 in the system log that states the application pool has requested a recycle because it has reached its virtual memory limit. The only conclusion I can come up with is that KVM manages memory differently than VMWARE. I did notice that the IIS application pools in VMWARE have a NUMA field that has a value of "maxavailablememory" in the CPU section. The IIS app pool in KVM do not have this field.
On the IIS VM that uses the KVM hypervisor, all of my applications have dedicated app pools and the only memory value that eliminates the issue is 0. We've tried increasing all of the memory values.
Has anyone ran into an issue like this before with a successful outcome, or have any suggestions that I could try to resolve the issue causing the 5077 events that are continuously posting?