Chart controls and notes
Showing the first 20 categories. Open full screen to explore all 95 categories.
Analyzing service-level spending trends over a time period can expose expected and unexpected growth in the consumption of various services across all providers and is a good way for the FinOps team to prioritize usage and rate optimization conversations with product, engineering and finance leaders within the organization.
This query uses the Provider column. Provider is deprecated in v1.3 and will be removed in v1.4.
Column names from the source use case.
Showing the first 20 categories. Open full screen to explore all 95 categories.
95 rows. Scroll inside the table to see more.
select date_trunc('month', charge_period_start at time zone 'UTC') as charge_month,
provider_name as provider_name,
service_name as service_name,
sum(effective_cost) as total_effective_cost
from focus.cost_and_usage
where charge_period_start >= $1::timestamptz
and charge_period_start < $2::timestamptz
group by date_trunc('month', charge_period_start at time zone 'UTC'), provider_name, service_name
order by charge_month asc, total_effective_cost desc nulls last, provider_name asc nulls last, service_name asc nulls last
limit 500SELECT MONTH(ChargePeriodStart), ProviderName, ServiceName, SUM(EffectiveCost) AS TotalEffectiveCost FROM focus_data_table WHERE ChargePeriodStart >= ? AND ChargePeriodStart < ? GROUP BY MONTH(ChargePeriodStart), ProviderName, ServiceName ORDER BY MONTH(ChargePeriodStart), SUM(EffectiveCost) DESC