The Performance Cost of Transparent Data Encryption in PostgreSQL

A technical whitepaper on pg_tde 2.2.2, including the numbers most vendors don’t show you.

Does encryption slow down PostgreSQL? Every vendor can produce a graph where transparent data encryption looks free, standard benchmarks never generate the workloads where PostgreSQL TDE overhead actually appears. This whitepaper runs a real PostgreSQL encryption performance benchmark on pg_tde 2.2.2, using hardware built to expose overhead at rest, not hide it, and publishes the worst case next to the typical case.

What’s inside

• Why relation file encryption sits inside measurement noise for standard OLTP traffic
• The four bulk-rewrite operations where PostgreSQL encryption overhead does show up, with real numbers
• The WAL encryption PostgreSQL locking mechanics behind high-concurrency overhead, and why it’s hard to avoid
• What pg_tde 2.2.2 fixed, and the architectural work closing the remaining gap
• Full benchmark methodology and hardware specs, reproducible against your own workload, so you can answer “is pg_tde encryption fast” for your own environment

Who it’s for

Database engineers and DBAs evaluating PostgreSQL encryption at rest under compliance pressure (GDPR, HIPAA, PCI-DSS, SOX) who need real pg_tde performance numbers, not a vendor’s single flattering graph.

Technical decision-makers comparing the best PostgreSQL encryption at rest solution and want to know where transparent data encryption performance cost actually lives before committing to Percona Distribution for PostgreSQL or any other implementation.