Network Latency measurements - NEW with Login PI 1.5

From Login VSI Documentation
Jump to: navigation, search

The Login PI Dashboard contains graphical output for network latency measurements that pertain to the target/endpoint machine's resources. These latency times refer to the actual round-trip time that the protocol takes to complete its journey from the launcher to the target machine and back again.

These measurements provide a very good indicator in relation to system responsiveness, which can help with maintaining SLAs, show trends in high resource consumption, and specifically show what might be causing problems on clients/endpoints.

Description

Screenshot

  • The y-axis shows latency and the x-axis shows time
  • Different time-frames can be selected on the bottom of the card
Average Latency over time high-level view

Hovering over points shows a granularized-metrics tooltip:

  • Latency: Endpoint networking latency
  • Outbound bitrate: Actual bitrate of the outbound transmission
  • CPU Load: CPU load (on server-side for endpoint delivery)
  • Memory Load: Endpoint memory load
  • Number of Processes: Number of running processes on target
Average Latency over time granularized metrics

Click on More... button to expose Protocol Latency per launcher and host. This can help point out potential problem hosts and launchers, or those that might need attention.

  • Vertical orange line segments (y-axis) on the bar graph show what the threshold is set to
    • This is also what the horizontal line segments denote (x-axis) on the adjacent line graph
  • The yellow line shows the latency of active hosts and launchers over time, juxtaposed against the orange threshold line
Protocol Latency per Launcher and Host

Clicking on the card's question mark in the upper-right will expose a "help" card.

Average Latency help card