App Responds Time - Palo Alto Networks
Prisma SD-WAN uses application response time to determine the path a flow may take and confirms that the path adheres to the application SLA. Application response time is a combination of Network Transfer Time (NTT), Round Trip Time (RTT), and Server Response Time SRT) metrics and calculated for a flow on each path before a decision is taken to send a flow on that path. These metrics reflect in the App Response Time chart. The chart displays detailed information on the application transaction time for each prefix within an application. It determines network and server performance for a specific application, including information from the moment the client generates a request to the time the server receives the response in the cloud or the data center. It also takes into account L1 – L3 and L4 – L7 characteristics of an application, including end-to-end performance rather than just latency, jitter, and packet drops for an application. App Response Time chart reports on a per-app basis. Select at least one app to view the App Response Time by Path table. The health of the selected paths are indicated by color. You can view specific Health Events table associated with each application instance or service with transaction or init failures. Refer to the above table to understand descriptions on the different health states. Using NTTn, RTT, SRT, and UDP-TRT metrics, this chart provides information on the source of an under performing application.
- Network Transfer Time (NTTn)—The measure of network congestion. The amount of time it takes to transfer incoming data from an external server to a local client.
- Round Trip Time (RTT)—The measure of network latency. RTT is measured only for TCP flows and defined as the time taken between a forward and return related protocol exchange; TCP SYN to SYN-ACK for outbound flows, TCP SYN-ACK to ACK for inbound flows and the time between a data sequence and ACK of that data sequence.
Thus, RTT is measured throughout the life of a flow and not just at the TCP establishment. Measuring RTT throughout the flows life allows the system to account for TCP proxy devices like WAN optimization in the path, providing a more accurate measurement of RTT.
- Server Response Time—The amount of time it takes for the server to start transmitting data after it has acknowledged the client’s request. SRT measured for TCP flows only from the time request is received to the time the server sends the first response packet.
- UDP Response Time (UDP-TRT)—The amount of time it takes for the server to respond to the UDP transaction request from the time the request is received. Currently, UDP-TRT provides information on UDP DNS traffic only.