I am contracting at mid size organisation, and there is one small site where both the DCs (Hyper-V VMs on Windows Server 2008R2 with 4GB RAM) are redlining their memory usage a few times of the week. To me the evidence continually suggests that it is not sufficiently specced in terms of memory, but UMDH does not seem to outline memory leaks. As is common with DCs the LSASS.exe process is the chief culprit, and I have suggested to the local site person that the memory on both boxes needs to be increased. However, this is not so easy as the Hyper-V hosts do not have any spare nor have any capacity left to increase the physical RAM, so he is stonewalling. He wants me to explain why it is redlining more frequently than it did a year ago though (apparently) the infrastructure / clients have not changed. I do not thing it is a splurge of authentication requests from the network, as Perfmon does not see massive activity on the network (“busiest network adapter is less than 15%”). Basically he wants a breakdown of HOW LSASS is using the server memory and for what. Is there are a relatively straight-forward method of getting this information? Thanks
↧
