Map/Reduce Yielding

A key advantage of the map/reduce script type is that it can manage some aspects of its own resource consumption. This behavior is achieved through a feature known as yielding.

To understand yielding, first be aware that all map/reduce scripts are processed by SuiteCloud Processors. A processor is a virtual unit of processing power that executes a job.

With yielding, the system waits until after the completion of each map or reduce function invocation, then checks to see whether the job has exceeded certain limits. If it has, job gracefully ends its execution, making it possible for other jobs to gain access to the processor. A new job is created to take the place of the map/reduce job that ended. The new job has the same priority as the old one, but a later timestamp.

Yielding can occur after the following limits are surpassed:

• The time limit specified by the Yield After Minutes field on the script deployment record. You can set this value to any number between 3 and 60. The default is 60. For more details on this field, see Yield After Minutes.

• The governance limit of 10,000 usage units per map or reduce job. For more details, see Soft Limits on Long-Running Map and Reduce Jobs.

Because the system checks these limits only between function invocations, yielding never interrupts a function invocation. Additionally, be aware that a job does not yield until it has passed one of the relevant limits.

Yielding is unrelated to the governance limits that exist for a single invocation of a map or reduce function. Those limits are described in Hard Limits on Function Invocations. Exceeding the limit for a single invocation of a map or reduce function causes the system to throw an SSS_USAGE_LIMIT_EXCEEDED error and ends the function invocation, even if it is not complete.

For help understanding how the map and reduce stages can each have multiple function invocations, review Map/Reduce Key Concepts.

General Notices