Skip to content

Queue Processing

When to Use

Use cron-based processing for standard setups. Use standalone endpoints when you need higher-frequency processing, custom scheduling (Jenkins, external cron), or separation from Drupal's cron run.

Decision

Approach Use When
Cron-based processing Standard setup, manageable volume
Standalone push endpoint Custom scheduling, high-frequency push
Standalone pull endpoint Custom scheduling, high-frequency pull
Per-mapping standalone flag Mix of cron and standalone for different mappings

Pattern

Cron processing limits: 

Push: salesforce.settings:global_push_limit (default 10,000/run)
      Per-mapping: salesforce_mapping.[id]:push_limit
      Failed retry: salesforce_mapping.[id]:push_retries

Pull: Populated by pull_frequency and pull_trigger_date
      Max queue: salesforce.settings:pull_max_queue_size

Standalone endpoint configuration: 

# Global (salesforce.settings.yml)
standalone: true

# Per-mapping
push_standalone: true
pull_standalone: true

Standalone endpoint URLs: - Push: /salesforce/push/process-standalone - Pull: /salesforce/pull/process-standalone

External scheduler setup: 1. Enable standalone: true in salesforce.settings.yml 2. Enable push_standalone or pull_standalone per mapping 3. Configure external scheduler (Jenkins, system cron, etc.) 4. Schedule HTTP requests to the standalone endpoint URLs

Common Mistakes

  • Wrong: Enabling standalone mode globally without also setting per-mapping standalone flags → Right: Global standalone enables the endpoints; per-mapping flags control which mappings use them
  • Wrong: Not monitoring salesforce_push_queue for items where fails > 0 → Right: Set up monitoring or periodic drush queue:list checks; failed items will not retry beyond push_retries

See Also