What is the function of this option Pre-aggregated statistics import in the VC Extractor Service ETL? Does it perform an aggregation that the Data Hub would do in order to improve Data Hub performance? Or is it doing some type of data summarization to the metrics prior to data load? What is the impact to having it turned on or off from both ETL Engine and Data Hub perspective?
What is the difference between the Saver Period and Hierarchy import period?
What is the difference in the two options for extracting Cluster/Pool performance data?
When CO import data, the statistic like max, min and average are computed (by the warehouse) starting from the hour resolution. So, at hour resolution, the real value of the sample and the statistics like min, max and average has all the same value. If you look for example at day resolution, the statistic are computed using the 24 hours samples. So, the max is the max value in the 24 hours.
Enabling the option in the ETL, it will import the statistic from the vCenter. So, at hour resolution, you will have the min, the max and the average that are the real statistics in the hour, returned by the vCenter from the raw data. In detail, what you will see will be the min and the max computed on the samples at 5 minutes of resolution. You are not losing data keeping the option off, but you will have different values if you perform an analysis that show the statistics.
There is not a "best choice" for this option: you have to set it according with your needs. There is not performance differences between the two options.
The ETL make a poll every 15 minutes. During this pool, it extract date from the vCenter and store this data into a local DB on the ETL Engine. Every hour or 2 hours (according with the value set in the property Saver Period) data are store into CO DB (stage area), using SQL Loader (https://bmcsites.force.com/casemgmt/sc_KnowledgeArticle?sfdcid=000029255)
The hierarchy import period is the interval for which the ETL import the information related to the relationship across the entity (e.g. this VM is under this cluster).