We are seeing quite a lot of problems generated with the message"TCP connectivity rate for process IIS on host xxxxxxxx has decreased to 33 %" (sometimes also 0%), but the service is up and running.
I can see this happening accross differents DT tenants.
Is there any reason why those problems are raised? Any way to fixe it?
This metric is related to the Percentage of properly established TCP connections compared to TCP connections that were refused or timed out. So there are either some timed out connections or refused connections for that process.
If you do not want to alerted for TCP connectivity problem of processes then you can disable this from anomaly setting of host but in this way it is going to be disabled for all the processes running on that host.
If you still want to be alerted for some critical processes then you can create the custom alert and set the threshold as per you need.