Response Time
Response time is the time from when a client of the service sends a request to when the client receives the response. This is the total of the network RTT plus the server delay excluding any client delay.
Response time is useful as an overall indicator that the responsiveness of an application has changed without pointing to a specific cause (i.e., client, server or network). It includes delays introduced by both the network and specific servers.
An increase in response time is an indication that overall application responsiveness has decreased. In order to obtain insight into this, you can determine whether the cause is the network (RTT - Network Round Trip Time) or a server issue (Server Delay). This can be determined by right-clicking the server outlier and requesting traffic flow information. Review the traffic flow details to determine which components of the metric have increased.
You can also consider how many servers in a service had an increase in response time. If ALL servers in a service had an increase, it could indicate a network problem.
Increased response time could be the result of:
-
Increases in Network RTT
-
-
Network congestion causing a general slowdown in connectivity
-
Packet losses
-
Routing changes
-
-
Increases because of Server Delay
-
-
High server load
-
Server application problems
-
Back-end dependencies having trouble
-
Monitor this metric to understand the Response Time for a service. By default this metric is enabled to alert you to spikes in response time. You can use the advanced settings if you are also interested in monitoring for dips in response time.