# AWS IoT Jobs limits
<a name="job-limits"></a>

AWS IoT Jobs has Service quotas, or limits, that correspond to the maximum number of service resources or operations for your AWS account.

**Topics**
+ [Job executions limits](#job-execution-limits)
+ [Active and concurrent job limits](#job-limits-active-concurrent)

## Job executions limits
<a name="job-execution-limits"></a>

This section provides information about the job execution limits for AWS IoT Device Management.

**Note**  
These limits are not part of the service quotas that you can find in the [AWS IoT Device Management Service Quotas documentation](https://docs.aws.amazon.com/general/latest/gr/iot_device_management.html#iot_device_management_quotas).

To get information about the number of pending job executions, you can either use the `GetPendingJobExecutions` API, or subscribe to the MQTT reserved topics for AWS IoT Jobs and receive [Job notification types](jobs-comm-notifications.md#jobs-comm-notifications-types).

The number of pending job executions in your account can vary depending on whether you have the scheduling configuration enabled and use a recurring maintenance window.


**Maximum number of pending job executions**  

| API/notification name | Description | Without scheduling configuration | With scheduling configuration | 
| --- | --- | --- | --- | 
| ListNotification | A ListNotification is published whenever an old job execution enters a terminal status, or when a new job execution is queued or changes to a non-terminal status. It can display up to 15 pending job executions that are either QUEUED or IN\$1PROGRESS. | 10 | 15 (Up to 5 job executions only appears in the ListNotification during a maintenace window). | 
| GetPendingJobExecutions | When you invoke the `GetPendingJobExecutions` API, it returns a list of job executions that have not yet started, and can be started after the API call. The API can return up to a maximum of 10 pending job executions. [\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/iot/latest/developerguide/job-limits.html)  | 10 | 15 | 

## Active and concurrent job limits
<a name="job-limits-active-concurrent"></a>

This section will help you learn more about active and concurrent jobs and the limits that apply to them.

**Active jobs and active job limit**  
When you create a job by using the AWS IoT console or the `CreateJob` API, the job status changes to `IN_PROGRESS`. All in-progress jobs are *active jobs* and count towards the active jobs limit. This includes jobs that are either rolling out new job executions, or jobs that are waiting for devices to complete their job executions. This limit applies to both continuous and snapshot jobs.

**Concurrent jobs and job concurrency limit**  
In-progress jobs that are either rolling out new job executions, or jobs that are canceling previously created job executions are *concurrent jobs* and count towards the job concurrency limit. AWS IoT Jobs can roll out and cancel job executions swiftly at a rate of 1000 devices per minute. Each job is `concurrent` and counts towards the job concurrency limit only for a short time. After the job executions have been rolled out or canceled, the job is no longer concurrent and does not count towards the job concurrency limit. You can use the job concurrency to create a large number of jobs while waiting for devices to complete the job execution.

**Note**  
If a job with the optional scheduling configuration and job document rollout scheduled to take place during a maintenance window reaches the selected `startTime` and you're at your maximum job concurrency limit, then that scheduled job will move to a status state of `CANCELED`.

To determine whether a job is concurrent, you can use the `IsConcurrent` property of a job from the AWS IoT console, or by using the `DescribeJob` or `ListJob` API. This limit applies to both continuous and snapshot jobs.

To view the active jobs and job concurrency limits and other AWS IoT Jobs quotas for your AWS account and to request a limit increase, see [AWS IoT Device Management endpoints and quotas](https://docs.aws.amazon.com/general/latest/gr/iot_device_management.html#job-limits) in the AWS General Reference.

The following diagram shows how the job concurrency applies to in-progress jobs and jobs that are being canceled.

![\[Image showing the different states of an AWS IoT job.\]](http://docs.aws.amazon.com/iot/latest/developerguide/images/job-states-concurrency.png)


**Note**  
New jobs with the optional `SchedulingConfig` will maintain an initial status state of `SCHEDULED` and update to `IN_PROGRESS` upon reaching the selected `startTime`. After the new job with the optional `SchedulingConfig` reaches the selected `startTime` and updates to `IN_PROGRESS`, it will count towards the active jobs limit and job concurrency limit. Jobs with a status state of `SCHEDULED` will count towards the active jobs limit, but will not count towards the job concurrency limit.

The following table shows the limits that apply to active and concurrent jobs and the concurrent and non-concurrent phases of the job states.


**Active and concurrent job limits**  
[\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/iot/latest/developerguide/job-limits.html)

**Note**  
The max duration of a recurring maintenance window is 23 hours, 50 minutes.