Proactive Alert Monitoring: Dequeue Latency

Salesforce is built on a multi-tenant structure and as such, dequeue times can vary depending on the current load and available resources. All Asynchronous processes are typically run in the background at a later time once resources become available and may be subject to dequeue latency. As asynchronous processes have no SLA, poorly implemented requests can cause performance issues if they become stuck in the queue or run very long.

A Dequeue Latency alert indicates that asynchronous jobs are taking a longer time than usual in the queue before processing by the server. A Dequeue Latency alert will be for one of the following asynchronous job types:

• Apex Future
• Apex Queueable
• Dashboard
• Bulk API

The MQ Longest Wait Time monitor will track messages that are still in flight but have not yet been dequeued, whilst the Dequeue Latency monitors will alert you when specific features (as listed above) have been dequeued through saw increased time to do so.

Please note there is no SLA for dequeue time.

The Signature Success plan’s Proactive Monitoring product offers multiple monitors to monitor for, and alert you to, sustained increases in dequeue latency or wait times. This article outlines some scenarios to help you investigate should you receive an alert. In addition, on the Salesforce Help site Designated contacts and above can see your realtime metrics, alerts & more.

Causes

Dequeue Latency may occur for a variety of reasons:

• Large volume of a particular job type in the queue
• Implementation of Fair Usage Policy
• Performance Degradation