Can I ask where the metrics in this graph are generated? because the teal part which I have though is total used by sql server, but it doesn't match resource mon, and it doesn't match sys.dm_os_process_memory.
A secondary question, is why would the top line be going down? I thought this represented the total physical available, and is non-changing. I've never seen this line change on any other server I monitor.
Here is another image from this same server. from a previous time
and when the teal area meets the top line, the server becomes unresponsive with errors that would seem to suggest memory exhaustion. Has anyone seen anything like this before?