Production Engines Versus Development Engines
Production Engine | Development Engine |
State information is needed for recovery: current state logging to the engine database is always turned on.
Backup engine unit and governor is needed for failover.
A number of database services are normally required to meet engine performance (database access) requirements.
State information does not need to be stored in the engine database: current state logging can be turned off.
Failover is not critical: a minimal engine configuration is adequate.
A single database service is normally adequate.
Historical process execution information is needed to analyze and improve system performance: logging of historical state information is normally turned on.
Engine is rarely cold started: registration and history information would be lost.
Application upgrades need to be performed without interrupting process execution.
Resource limitations need to be monitored to ensure engine performance and stability: memory, disk space, and cpu utilization.
Current process execution information is more important than historical information: history logging can be turned off.
Engine is often cold started: registration information is purged, allowing for a clean engine state.
Engine can be shut down and cold started to perform application upgrades.
Resource limitations rarely are of concern in development environments.
Keep these issues in mind when configuring an engine. In most situations, you have to configure and manage at least one development engine and one production engine.
NOTE Do not use the same engine for both development and production.
