Capacity Detail Modal
Last updated
Last updated
The capacity detail screen shows a heat map with 3 months of calendar data.
You can see the condition of the selected router and circuit at a glance.
For each day, the most severe or highest saturation, a single 10-minute time period, is what informs the calendar’s visual display. On the calendar, up arrows are outbound, down arrows are inbound. For example, if January 5th shows deep red with a up arrow, you’d click it and then look in the Egress timeline below for the tallest of the deep red bars.
Any day on the calendar that’s visually flagged with yellow or red is a potential problem – but you won’t know until you consider the context and view the detail timeline. A single red day could indicate a day that was saturated all the time, or maybe not. To investigate, click on a day to see whether there were a lot of saturated periods or just one anomalous spike.
Another clue that you are approaching saturation is if you see a lot of red or yellow days in a row, or consistently over many days or even months.
The timeline displays for the calendar date selected. Each vertical bar below is a single 10-minute time increment within the 24-hour period selected above. To change office hours settings or warning thresholds from this screen, click Capacity List at the top and then click Settings.
Colored bars indicate activity within normal business hours. Pale gray bars indicate times of day that are outside of normal business hours. This is if you have Office hours only selected in Settings. Otherwise, all bars are colored.
A heatmap legend at the top matches the user-defined warning thresholds for Moderate to Severe. For example if you’ve set a 60% saturation level as a Moderate warning level, then every 10-minute increment that has 60% saturation will show as yellow, whereas if you set Moderate to 85%, those same increments would be green.
When the detail heatmap shades from red to black instead of yellow to red, the black color indicates utilization well beyond actual capacity. Since it’s highly improbable to see 1000% utilization, most likely this means that the capacity configuration is incorrect and does not accurately reflect the real-world capacity. If you see a screen like the one below, verify your capacity settings first.
Top talkers refers to individual applications, rather than application categories as shown on the Recommendations overview or Recommendation Detail screens. For example, you might see something like ms-teams for Microsoft Teams, even if Microsoft Teams is bundled into a larger category. These applications are customizable in vManage.
When you click a 10-minute increment in one of the timelines below, Top Talkers displays, and refers to the percentage of capacity used by each application category, within that 10-minute time period. If there was a spike in utilization, which applications caused it? Is that consistent throughout the day? Each of the graphs adds up to 100% utilization.
The guiding principle behind capacity planning is to quickly surface potential interface saturations, determine if it’s really saturated – and, if it is a persistent pattern of saturation, what to do about it.
So, you went to the list page and saw some red numbers on the calendar history. You still have to figure out why that particular router and circuit is saturated, and which applications are the “top talkers”. For example:
Are more people coming to the office?
Is there a new weekly onsite meeting causing a spike?
Is the traffic from applications mainly designed for personal entertainment?
Was there a surge caused by malware or some sort of malicious attack?
Is this pattern repeated across multiple sites and interfaces?
Do the egress and ingress saturations seem to match up, or is there an imbalance?
Next go to the capacity details page, and review the 3-month calendar by day. Then start clicking one day at a time in the calendar to view the 24-hour time series for that day, in 10-minute intervals.
Typically, organizations will pay attention to a circuit as it enters the yellow or warning state. If the utilization levels seem consistent, and patterns indicate that it is likely to increase, there will be enough lead time to negotiate an increase in bandwidth (capacity) with your ISP, a process which can take several months.
Finally, you can investigate path quality on the Site Details Screen. Note that the Site Details screen is application-specific (application category or class). So, if you see high utilization on the Capacity screens, you can visit Site Details to compare the quality for an application category such as Voice, and see if any of those Voice applications show up as Top Talkers on the Capacity screens.
Is there a correlation between capacity saturation and path quality?
What about the number of active users?
Is the capacity data correct? If everything else looks good, make sure you’re not showing false positives.
