There is often confusion around auto‑scaling in Pega Cloud, especially when running large batch jobs such as Data Flows or migrations. Here’s a simple clarification based on real production behavior.
What does auto‑scale in Pega Cloud
Pega Cloud auto‑scales primarily for real‑time, user‑driven workloads, including:
-
User / portal traffic
-
REST and API requests
-
External integrations and channels
Scaling is metric‑based (CPU, memory, concurrency) and happens within pre‑configured limits to protect system stability.
What does NOT auto‑scale
Pega Cloud does not auto‑scale batch workloads, such as:
-
Batch Data Flows
-
Large data migrations
-
ETL‑style processing
-
One‑time or scheduled backfills
If a production system has 2 batch nodes, batch jobs will run only on those nodes—additional nodes are not added automatically, even for large workloads.
