About the way 95 values are calculated in csv datasheet?

[aliuzh97]

Nth Percentile |95:bits:0:aggregate_current_peak:2|
Nth Percentile |95:bits:0:aggregate_sum_peak:2|
Nth Percentile |95:bits:0:aggregate:2|


Nth Percentile|95:bits:0:aggregate_max:2|

With the 0.8.8 I'm using, the 95 values are calculated using aggregate_max so that the data obtained is consistent with the traffic data provided by my client, which I think is accurate

But after I changed to 1.2.27, I can't find the “aggregate_max”, cacti defaults to "aggregate_sum_peak", which is inaccurate, and I can't find the 95 value in the csv, including "aggregate_current_peak" and "aggregate:2", I want to know how to modify the template or aggregation, so that the 95 value of the csv table is aggregate_max.

[TheWitness]

I think the calculations are done in graph_variables.php in the library directory.

I would be open to suggestions. I know that especially for aggregate graphs it's not what it needs to be.

[aliuzh97]

I've reviewed the calculation method in 'graph_variables.php'. However,currently my requirement is: I want only the value defined by cacti as 'aggregate_max' to be treated as the 95th percentile value. It is most accurate for my purposes, and I'd like it to be displayed in both Cacti graphs and CSV tables.And The traffic values obtained in this manner are the most accurate, allowing me to use these figures for financial reconciliation with others. This ensures that all parties have a consistent and precise understanding of the usage, which is crucial for billing and cost management purposes. By using the 'aggregate_max' value as the 95th percentile, we maintain the highest level of accuracy in our calculations.

As for further suggestions, please consider implementing the 95th percentile value as an optional display setting. This way, users can choose which calculation method they prefer for their final results."

I really hope for a detailed response here. I've been a dedicated user of Cacti since version 0.8.8 through to 1.2.27.Given my long-standing experience with Cacti, I'm confident that implementing the 'aggregate_max' value in aggregate graphs as the 95th percentile will not only enhance the accuracy of the data but also align closely with best practices in network traffic analysis.

[aliuzh97]

I think the calculations are done in graph_variables.php in the library directory.

I would be open to suggestions. I know that especially for aggregate graphs it's not what it needs to be.

I want to get a relatively accurate answer, and then I test it, which is something that is very urgent for me

[aliuzh97]

Anyone can answer this question？