SQL hosting is the practice of provisioning, operating, and scaling a relational database (PostgreSQL, MySQL, SQL Server, Oracle, etc.) on external infrastructure managed by a cloud or hosting provider. Engineers interact with the database using standard SQL and connection strings, while the provider handles hardware, operating system, patching, backups, high availability, monitoring, and automatic scaling.Key points:- Not a reserved word in any SQL dialect; appears in documentation and marketing as shorthand for Database-as-a-Service (DBaaS).- Eliminates the need for sysadmin work such as RAID configuration, kernel tuning, and physical backups.- Costs are billed per usage (compute hours, storage, I/O). Idle databases may still incur storage charges.- Built-in redundancy and failover improve reliability but can increase latency compared to on-prem deployments.- Vendor-specific features (extensions, monitoring dashboards, proprietary roles) may create lock-in.Caveats:- Limited superuser access can block low-level optimizations.- Network latency and egress fees can surprise teams moving large data sets.- Data residency and compliance requirements must be verified with the provider.
DBaaS, CREATE SERVER, DATABASE LINK, FOREIGN DATA WRAPPER, CONNECTION STRING
N/A - infrastructure concept, not in the ANSI SQL standard
A managed service model where a provider runs the database engine for you. You only supply SQL and connection credentials.
No. Queries remain identical. Only administrative commands such as creating users or extensions may require provider-specific tooling.
Most vendors charge for provisioned compute cores, storage, backups, and I/O operations. Some offer serverless per-second billing.
Yes, export a logical dump (e.g., pg_dump, mysqldump) or use replication. Watch for proprietary features that lack on-prem equivalents.