QUESTION:
What is the Traverse historical performance data aggregation scheme?
RESOLUTION:
Traverse stores historical performance data for 3 years using the following progressive aggregation scheme that automatically reduces the granularity as the collected data ages:
24 hours of 5 minute samples
1 week of 15 minute samples
3 months of 1 hour samples
3 years of 1 day samples
That is; each (aggregated) data point records the maximum, minimum and average for the samples collected during the aggregation interval (i.e. 5 min, 15 min, 1 hour, 1 day).
This means that for a test with a polling interval of 5 minutes, each data point in the most granular scheme (5 minute samples) will be composed of a single sample. For a test with a polling interval of 1 minute, each data point in this same scheme represents 5 polled values.
Note:
- On Premise customers: Upon request, it may be possible to change the default aggregation schemes at the time of installation. Please open a support request where this may be discussed further. We generally advise customers not to deviate from the default aggregation schemes unless absolutely necessary, as this may affect the Traverse performance and storage requirement.
- SaaS customers: It is not possible to change the default aggregation scheme
APPLIES TO:
All versions of Traverse