Data gravity is the tendency for the environment containing most of an organization's data to attract its primary applications, services, and infrastructure. Moving data between on-premises systems and cloud providers, or between clouds, consumes time and money through transfer costs. section 5.6.1
In observability, data gravity makes backend placement a long-term decision. Ingestion, storage, and query costs accumulate, while transferring signals between separate backends can create substantial export–import costs. Different signal types may require different backends, so interoperability matters. section 5.6.1 section 5.7
Use data gravity when selecting backend environments:
- Place data where its dominant environment minimizes unnecessary transfer between on-premises and cloud or between cloud providers. section 5.6.1
- Compare ingestion, retention, storage, and query costs rather than evaluating storage price alone. section 5.6.1
- Prefer open query languages and wire formats to reduce lock-in when signals use multiple backends. section 5.7
