Across engines

View as Markdown

Datapace is engine-agnostic by design. The signals it reads differ per engine; the graph, the agents, the policy, and the ledger are the same for all.

Engine-agnostic by design

Mid-size and large estates run several engines. The sales system is on Postgres since 2016, finance came with an acquisition on SQL Server, the inherited Db2 nobody dares touch is from 2011. A tool that covers one of them leaves the team with the same problem on the other two.

Datapace is designed for the whole estate. One graph spans every engine, and an entity can resolve to tables in more than one of them.

What is engine-specific

The signals. Each engine exposes its schema, its workload, and its statistics in its own way, and Datapace reads each one in the way that engine offers. A query-statistics view on Postgres, a plan cache on SQL Server, a catalog on Db2: different sources, the same facts once they land on the graph.

The infrastructure facts differ too: a managed cloud service reports instance class and cost one way, a self-hosted engine another. Both land on the graph as instance class, storage, cost, performance, usage, and freshness.

What is not

Everything above the signals:

  • The graph. Entities, measures, relationships, semantics, lineage: one model, whichever engine holds the tables.
  • Infer and confirm. The same scores, the same states, the same working sessions.
  • The agents. FinOps, Performance, Quality, Migration, and Documentation draft on the graph, not on the engine.
  • Policy, approval, and the ledger. One set of rules and one record for the estate.
  • Governed context. Copilots, BI, and MCP clients see one graph, not one per engine.

For a managed-service provider this is the point: one way of working across every client's engines, with each client's estate on its own graph.

Engines

Datapace is designed for the engines an estate actually runs. The families named on the site are the relational engines (PostgreSQL and its managed forms such as Amazon Aurora, Google Cloud SQL, Neon, Supabase, and Timescale; MySQL and MariaDB; Microsoft SQL Server; Oracle; IBM Db2; CockroachDB), the document and key-value stores (MongoDB, Amazon DynamoDB, Azure Cosmos DB, Redis, Apache Cassandra), and the analytical engines (Snowflake, ClickHouse, Elasticsearch).

Which engines and which environments Datapace reads in your estate is part of the scope agreed with your team.

Was this page helpful?