tencent cloud

All product documents
Elastic MapReduce
Billing Overview
Last updated: 2025-01-24 09:24:25
Billing Overview
Last updated: 2025-01-24 09:24:25

Billing Mode

EMR clusters support two billing modes: monthly subscription and pay-as-you-go. The cluster fees are the sum of fees for all nodes in the cluster and other involved cloud products. Elastic nodes (Task nodes) support the spot instance mode.
Nodes in monthly subscription clusters support the monthly subscription, pay-as-you-go, and spot instance mode. Nodes in pay-as-you-go clusters support the pay-as-you-go and spot instance modes.
The table below shows the differences among the three billing modes:
Billing Mode
Monthly Subscription
Pay-As-You-Go
Spot Instance
Payment method
Prepaid
Postpaid; amount freezing upon purchase, billed hourly
Postpaid; amount freezing upon purchase, billed hourly
Billing unit
USD/month
USD/second
USD/second
Unit price
Relatively lower
Relatively higher
The price fluctuates. In most cases, the price is about 10-20% of the price of a pay-as-you-go instance with the same specifications.
Minimum use time
At least one month (calendar month)
Charged by second and billed by hour. Purchase and release at any time.
Charged by second and billed by hour. Purchase and release at any time. May be repossessed by the system.
Configuration change
Upgrade configuration at any time. No limit for each node (CPU and memory only).
No limit. Change (node CPU and memory only) at any time.
Not supported
Use case
Suitable for long-term clusters with stable needs for compute resources.
Suitable for a cluster to exist for a short period or periodically.
Suitable for a cluster using elastic compute resources to get more computing power.
Billing mode change
Not supported
Not supported
Not supported
The pay-as-you-go mode offers 3-tiered pricing, except for model S5 and new models launched after November 2019. A longer usage period means a bigger discount.
Tier 1: 0 <T1 ≤ 96
Tier 2: 96 <T2 ≤ 360
Tier 3: T3 > 360 For more information, see Billing Overview.
You can view node specifications and pricing and estimate your resource costs on the Pricing | Elastic MapReduce page.
EMR bills only show the costs of nodes, namely, the costs of CPU, memory, system disk, and local data disk. The bills of associated cloud products are available in their respective consoles. For EMR bill details, see Viewing Bills or Cost Allocation by Tag.
Caution
Select the shutdown mode with caution when shutting down a pay-as-you-go EMR cluster node in the CVM console, because EMR nodes do not support the "no charges when shut down" mode.
The prices of a model given on the Pricing | Elastic MapReduce page cover only the configuration of its CPU, memory, and local data disk, excluding the costs of images, cloud system disks, cloud data disks, and associated cloud products.
For the pricing of different types of disks, see Price Overview. For charges of associated cloud products, see "Billable Items > Charges of associated cloud products" below.
The prices are subject to changes as appropriate. Please visit our official website for the latest prices.

Billable Items

The cost of a cluster is the sum of the costs of all nodes in the cluster and of associated cloud products. Cloud products such as Elastic IP, CDB, CBS, CHDFS, COS, and Virtual Private Cloud (VPC) may be used when you use EMR. You will be charged for such products in their respective billing mode. For details, see their respective billing document.
Cost Category
Billable Item
Resource Use in EMR
Billing Description
Pricing
Costs of EMR nodes
Node cost
You can select the model specifications as needed. The nodes are subject to separate pricing by EMR. The price of a node covers its resources such as CPU, memory, system disk, and local data disk. Three billing modes are available: pay-as-you-go ,monthly subscription and spot instance.
The EMR cost is only the costs of all nodes.
Costs of associated cloud products
Elastic IP (EIP)
Public network access is enabled for the Master.1 node in a cluster by default so that you can access the WebUI pages of various Hadoop components from outside the cluster. Traffic fees are incurred for data interactions when you visit these pages, but a little traffic is generated in most cases. Therefore, the bill-by-traffic mode is used by default, which costs less than the bill-by-bandwidth mode.
TencentDB for MySQL (CDB)
If you plan to deploy one or more components among Hive (local), Hue, Ranger, Oozie, Druid, and Superset in an EMR cluster, you need to purchase a CDB instance for storing metadata.
Cloud Block Storage (CBS)
If you plan to deploy cloud disks as the data disks in a node, you need to purchase at least one cloud disk. The cost of the cloud disk will be paid together with the EMR node purchased, with the corresponding CBS order generated. The cloud disk will be subject to the same billing mode as the node.
Cloud Object Storage (COS)
If you use COS to separate the compute and storage resources of a cluster, you will be charged for data storage and requests when the cluster requests to pull data stored in COS for computing. Meanwhile, result data storage, backup, or other operations in the computing will also generate new data in COS.
Cloud HDFS (CHDFS)
If you use CHDFS to separate the compute and storage resources of a cluster, you will be charged for data storage and requests when the cluster requests to pull data stored in CHDFS for computing. Meanwhile, result data storage, backup, or other operations in the computing will also generate new data in CHDFS.

Was this page helpful?
You can also Contact Sales or Submit a Ticket for help.
Yes
No

Feedback

Contact Us

Contact our sales team or business advisors to help your business.

Technical Support

Open a ticket if you're looking for further assistance. Our Ticket is 7x24 available.

7x24 Phone Support
Hong Kong, China
+852 800 906 020 (Toll Free)
United States
+1 844 606 0804 (Toll Free)
United Kingdom
+44 808 196 4551 (Toll Free)
Canada
+1 888 605 7930 (Toll Free)
Australia
+61 1300 986 386 (Toll Free)
EdgeOne hotline
+852 300 80699
More local hotlines coming soon