Unit Economics
A lower unit cost can coexist with a higher bill. Define the cost, output, period, and service conditions before interpreting the change.
-
-
Data Gravity
Data placement affects transfer and processing costs. Compare local execution, replicas, and remote reads using measured volume, lag, and workload needs.
-
Cost Allocation Tag
Tags, account ownership, and usage records answer different allocation questions. Measure missing coverage and reconcile the result against billing.
-
Data Transfer Cost
Map the actual flows and price each path. Assess replication, batching, and placement against correctness, latency, recovery, and availability requirements.
-
Dynamic Resource Allocation
Returning executors can reduce contention. Billing savings depend on node scaling and pricing; shuffle files and cached blocks affect when executors can leave.
-
Spark Performance and Cost Factors
More executors help while tasks await slots. Skew, layout, memory, and serialization can set limits that extra machines do not remove.
-
Capacity Unit (Microsoft Fabric)
An F64 capacity provides 64 capacity units, shared by every workspace assigned to it. The number sets compute, licensing rules, and who slows down whom.
-
Snowflake Credit
A Snowflake credit measures compute usage; the bill multiplies it by a price that varies by account and region. Storage and data transfer are billed otherwise.
-
OpEx
OpEx concerns expense recognition. Payment timing, asset depreciation, and minimum capacity commitments must be assessed separately.
-
Storage Tiering, Lifecycle, and Archival
Choose storage tiers from access and recovery needs, lifecycle filters, version retention, holds, and restoration checks.