Pay-as-you-go is suitable for projects that require flexible storage space and have fluctuations in the number of jobs along with time periods. This topic describes resources and billing rules of pay-as-you-go services. This topic also describes the limits and usage notes of pay-as-you-go services.
Resources
Basic resource | Description | Billing rule |
Computing resources | Computing resources are shared. These resources are consumed to run computing jobs, such as SQL jobs that involve user-defined functions (UDFs), MapReduce jobs, Spark jobs, Lightning jobs, and Graph jobs. High-priority computing jobs can preempt resources from low-priority jobs if resources are insufficient. You cannot specify or limit the amount of resources that computing jobs can consume. Note If a large number of jobs run at the same time, the existing resources may be insufficient for jobs due to multi-user preemption. As a result, the response latency occurs. | You are charged based on the computing resources that are consumed by different types of computing jobs. For more information, see Computing pricing. |
Storage resources | You are charged only for the resources that are used to store tables. MaxCompute compresses your data for storage. You are charged based on the amount of compressed data. In most cases, data is compressed to about 20% of the original data amount. For more information, see Storage pricing (pay-as-you-go). | |
Resources for data uploads | You are not charged for the resources that are used to upload data to MaxCompute. | |
Resources for data downloads | You are charged only for the amount of data that is downloaded over the Internet. For more information, see Download pricing (pay-as-you-go). |
Usage notes
For more information about how to activate pay-as-you-go services, see Activate MaxCompute and DataWorks.
After you activate pay-as-you-go services, you must take note of the following points:
For projects created in the region where the pay-as-you-go services are activated, you can select pay-as-you-go computing resources.
When you use pay-as-you-go services, the number of jobs is not limited. The computing resources that you use to run jobs are shared in a resource pool.
For example, a job requires 1,200 compute units (CUs). If the resource pool has sufficient resources, the job can consume 1,200 CUs. If all the resources in the resource pool are consumed or the number of available CUs is less than 1,200, the job must wait for resource release or preempt the remaining resources. As a result, the job execution latency occurs.
For more information about the switching between pay-as-you-go and other billing methods, see Switch between billing methods.
References
If you want to reduce the cost of running a job that is not time-sensitive during the analysis of large amounts of low-value data, such as data in user behavior logs or system logs, you can activate the Pay-as-you-go Spot Edition.