Internap | Intelligent IT Infrastructure solutions that deliver unmatched performance and platform flexibility

Enterprise IP

Intelligent IT Infrastructure solutions that deliver unmatched performance and platform flexibility

Route Performance Monitor Methodology

Methodology & Data Collection Notes

Managed Internet Route Optimizer™ (MIRO), Internap’s proprietary routing algorithm, collects volumes of data on Internet performance from our Private Network Access Points (P-NAP®) around the world. At any one location, MIRO makes hundreds of thousands of routing adjustments every day. By monitoring performance over all available carrier backbones and actively managing routes, MIRO ensures that traffic is sent over the best performing network at any given time.

RPM displays data pulled from MIRO to provide a view to both the overall performance of the Internet as well MIRO’s routing capabilities. Because RPM can only capture a tiny fraction of the MIRO algorithm’s routing probes and calibrations, the data shown in both the Compare Carriers and Monitor Internet tabs should be considered illustrative.

Monitor Internet

The best public Internet paths are determined by running trace routes between the selected P-NAP and destination points. After mapping IP addresses of each hop into a mapping database to obtain city names, and longitude/latitude information, RPM determines the worst round trip time (RTT) to each probed prefix (this is done at every P-NAP). Each of the suboptimal RTT readings is then associated with a specific carrier. Next, all other carrier routes probed the same prefix within +/- 3 minutes of the worst performing carrier are displayed.

RPM only return events with latency of 500 milliseconds (ms) or greater than the best carrier route to filter for unusual, carrier-specific latency events. RPM then selects 50 of these events to display for each 24-hour period. Each of these events is represented as a dot on the map. Depending on the severity (i.e., the difference between the best carrier route and worst carrier route) of the event, the dot may appear as a red halo (>=1,750 ms), red (between 800 and 1749 ms), orange (between 600 and 799 ms) or yellow (500 to 599 ms) color. By clicking on an event dot, a pop-up window appears displaying information each carriers’ latency performance for the selected location. Data values are refreshed every 24 hours and displayed on a rolling 60 day window.

Compare Carriers

Carrier latency values for are determined by a scheduled daily process that reads the probed values (collected using ping and trace route) stored in the performance table of the MIRO database. The data values on the chart are measured in milliseconds (ms). Carrier latencies are calculated by averaging suboptimal routes (higher latency, jitter, packet loss etc. as defined by MIRO algorithm) across all probed prefixes for each carrier at each location over a 24-hour period. Internap daily latency values are calculated by averaging optimal routes (lower latency, jitter, packet loss etc. as defined by MIRO algorithm) of all carriers across all prefixes within each 24-hour period. Calculation for average latency across carriers is defined as [sum of carrier latencies, excluding Internap] / [number of carriers]. To the left side of the “Compare Carriers” graph, latency difference is defined as [carrier with highest latency] – [Internap latency]. Graph data values are refreshed every 24 hours and displayed on a rolling 60-day window.

Legends:
Latency Difference is defined as the performance difference between Internap and the highest latency carrier used in the comparison. Latency difference= [carrier with highest latency] – [Internap latency].

The source presents a list of all the origin cities (P-NAP locations) monitored by RPM.

The control allows for RPM play/pause for the time series displayed. The demo will auto play from the beginning of the date range to the end.

Red columns mark a moment in time when there is a prefix route with large differences (>=1,750 ms) between the best carrier route(s) and the worst carrier route. These columns correspond with pulsing red dots on the map which also indicate a particularly high difference between best carrier route(s) and worst carrier route to a particular destination.

When a latency-event dot is clicked, a modal box pops up with more details on the latency event. Red = highest-latency carrier. Yellow = lowest latency carrier. The event dot located in the upper left hand corner corresponds with the severity bands described in the legend at the bottom of the map.