Talend Jobs
A Talend job is a packaged ETL process designed and exported from Talend Studio.
In EAC, these jobs are deployed as executable archives. Once deployed, jobs can be:
- Launched manually
- Scheduled for automated execution
- Monitored for performance and errors
EAC allows multiple versions of the same job to coexist, so you can test new versions without affecting production. See Testing job version’s.
Tasks (Manual & Scheduled)
A task in EAC represents the execution configuration for a specific Talend job.
Each task can:
- Use a specific job version
- Have customized context parameters
- Be run manually or via scheduler
- Send notifications on success or failure
- Require acknowledgment before restarting if an error occurs
Multiple tasks can point to the same job but have different execution settings. See good use of contexts.
Contexts & Environments
In Talend, contexts store variables (e.g., database credentials, file paths, API keys).
EAC enhances this by allowing:
- Common contexts per project (shared across multiple tasks)
- Per-task context overrides for customization
- Environment contexts (e.g., Development, Staging, Production) for clean separation
This ensures you can run the same job in different environments without modifying the job itself. See good use of contexts.
Multi-Server Mode
EAC supports multi-server execution for scalability and redundancy.
In this mode:
- Jobs can be deployed to multiple servers.
- Execution load can be distributed.
- Deployment synchronization.
- Each server instance must have a unique
APP_SERVER_NAME.
Logs & Purge System
EAC tracks and stores logs for every ETL execution, including:
- Start/end time
- Duration
- Execution status
- Detailed processing logs
- Metrics and statistics
Since logs can grow quickly in size, EAC provides four purge strategies:
- By Date – remove logs older than a given date.
- By Execution Count – keep only the last N executions.
- By Duration – remove executions shorter than X seconds (if successful).
- Manual or Scheduled – purges can be run manually or scheduled via the UI.
This ensures performance remains optimal without sacrificing troubleshooting capabilities. See good use of purges.
