All Products
Search
Document Center

E-MapReduce:Resource plans

Last Updated:Dec 06, 2025

EMR Serverless Spark offers resource plans that provide a more flexible billing model. These plans deduct the cost of computing resources from your pay-as-you-go bills. This topic describes the details of these resource plans.

Usage notes

  • Resource plans apply only to fees generated by pay-as-you-go queues in Serverless Spark workspaces. After you purchase a resource plan, the pay-as-you-go fees for all workspaces in the same region are deducted from the plan first.

  • Resource plans cannot cover pay-as-you-go usage incurred before you purchase the plan.

  • Usage that exceeds the resource plan quota or occurs after the plan expires is billed on a pay-as-you-go basis.

  • Resource plans cannot be renewed, upgraded, or downgraded.

Purchase and pricing

Purchase a resource plan

Note

You can purchase and stack multiple resource plans. If you have multiple plans, the plan that expires the soonest is used first by default.

  1. Go to the EMR Serverless Spark workspace page.

    1. Log on to the E-MapReduce console.

    2. In the navigation pane on the left, choose EMR Serverless > Spark.

  2. On the Spark page, click Purchase Resource Plan.

  3. Select the required parameters, such as Region, Resource Plan Capacity, Validity Period, and Purchase Quantity. Then, click Buy Now.

  4. Follow the on-screen instructions to complete the purchase.

Pricing and discounts

Important
  • For more information about pay-as-you-go prices in different regions, see Pay-as-you-go.

  • The discounted price can be combined with other business discounts.

  • If there are price discrepancies due to other promotions, the price on the purchase page prevails.

  • China regions

    Region

    Compute resource plan specification

    List Price (USD)

    Discount

    Discounted Price (USD)

    Plan validity (months)

    China (Beijing)

    China (Shanghai)

    China (Hangzhou)

    China (Shenzhen)

    China (Chengdu)

    3,000

    150.91

    95%

    143.36

    1

    5,000

    251.51

    90%

    226.36

    1

    10,000

    503.02

    85%

    427.57

    1

    50,000

    2,515.10

    80%

    2,012.08

    1

    100,000

    5,030.20

    80%

    4,024.16

    3

    China (Hong Kong)

    3,000

    222.14

    95%

    211.04

    1

    5,000

    370.24

    90%

    333.22

    1

    10,000

    740.48

    85%

    629.41

    1

    50,000

    3,702.40

    80%

    2,961.92

    1

    100,000

    7,404.80

    80%

    5,923.84

    3

  • Other countries and regions

    Region

    Compute resource plan specification

    List Price (USD)

    Discount

    Discounted Price (USD)

    Plan validity (months)

    Indonesia (Jakarta)

    Singapore

    3,000

    201.32

    95%

    191.25

    1

    5,000

    335.53

    90%

    301.98

    1

    10,000

    671.06

    85%

    570.40

    1

    50,000

    3,355.30

    80%

    2,684.24

    1

    100,000

    6,710.60

    80%

    5,368.48

    3

    Germany (Frankfurt)

    3,000

    194.38

    95%

    184.66

    1

    5,000

    323.96

    90%

    291.56

    1

    10,000

    647.92

    85%

    550.73

    1

    50,000

    3,239.60

    80%

    2,591.68

    1

    100,000

    6,479.20

    80%

    5,183.36

    3

    US (Virginia, Silicon Valley)

    3,000

    161.40

    95%

    153.33

    1

    5,000

    269.01

    90%

    242.10

    1

    10,000

    538.01

    85%

    457.31

    1

    50,000

    2,690.05

    80%

    2,152.04

    1

    100,000

    5,380.10

    80%

    4,304.08

    3

    Japan (Tokyo)

    3,000

    208.26

    95%

    197.85

    1

    5,000

    347.10

    90%

    312.39

    1

    10,000

    694.20

    85%

    590.07

    1

    50,000

    3,471.00

    80%

    2,776.80

    1

    100,000

    6,942.00

    80%

    5,553.60

    3

Example of how to select a specification

  • Estimate based on cost

    Assume that you have a Serverless Spark workspace in the Singapore region and your monthly pay-as-you-go cost is about USD 2,500. In this case, you can consider purchasing resource plans as follows:

    • Convert the monthly cost into computing resource usage in CU·H. The formula is: Total monthly cost / Pay-as-you-go unit price = USD 2,500 / (USD 0.067106/CU/hour) = 37,261 CU·H.

    • Considering the validity period of the resource plans, a conservative option is to purchase three 10,000 CU·H and one 5,000 CU·H resource plans.

    • After you purchase the resource plans, the monthly cost is:

      • Resource plan cost: Cost of three 10,000 CU·H plans and one 5,000 CU·H plan = (3 × 570.4) + (1 × 301.98) = USD 2,013.18.

      • Pay-as-you-go cost: Remaining resources billed on a pay-as-you-go basis = 2,261 × 0.067106 = USD 151.670.

      • Total cost: USD 2,013.18 + USD 151.67 = USD 2,164.85.

    • By choosing resource plans, you can save: USD 2,500 - USD 2,164.85 = USD 335.15.

  • Estimate based on computing resource consumption

    Assume that you have a Serverless Spark workspace in the Singapore region. Your cumulative computing resource consumption over the last 30 days is 25,500 CU·H. You can retrieve this information using the Resource Monitoring feature or by calling the GetCuHours operation. Assuming your monthly consumption remains similar, you can consider purchasing resource plans as follows:

    • Considering the validity period, a conservative option is to purchase two 10,000 CU·H resource plans and one 5,000 CU·H resource plan.

    • After you purchase the resource plans, the monthly cost is:

      • Resource plan cost: Cost of two 10,000 CU·H plans + Cost of one 5,000 CU·H plan = (2 × 570.4) + (1 × 301.98) = USD 1,442.78.

      • Pay-as-you-go cost: Remaining resources billed on a pay-as-you-go basis = 500 × 0.067106 = USD 33.553.

      • Total cost: USD 1,442.78 + USD 33.553 = USD 1,476.333.

    • Monthly cost before purchasing resource plans: 25,500 × 0.067106 = USD 1,711.203.

    • By choosing resource plans, you can save: USD 1,711.203 - USD 1,476.333 = USD 234.870.

View usage and set alerts

View usage details

After you purchase a resource plan, you can view its usage details in Expenses and Costs to help you manage the plan effectively.

  1. Go to the Resource Instances page in the Alibaba Cloud Management Console under Expenses and Costs.

  2. Set Resource Type to Resource Packages. You can then view the statistics for resource plans whose instance names contain emapreduce_sparkpostcu_dp. The statistics include the plan status, remaining quota, and validity period.

  3. Click the Details tab to view the detailed deduction records for the resource plan.

Set quota alerts

You can set quota alerts for your resource plans. An alert is triggered when the remaining quota of a plan falls below a specified threshold.

  1. Go to the Resource Instances page in the Alibaba Cloud Management Console under Expenses and Costs.

  2. Set Resource Type to Resource Packages. On the Instances tab, click Set Remaining Quota Alert in the upper-right corner. Set the remaining quota threshold for the alert and then click OK.

Set expiration notifications

After you set a quota alert for a resource plan, Alibaba Cloud sends you a notification by internal message, email, or text message when the quota reaches the threshold.

  1. In the top menu bar of the Alibaba Cloud Management Console, click the bt_console_message.png icon. In the navigation pane on the left, choose Message Settings > Common Settings.

  2. On the Product Message tab of the Common Settings page, find Notifications of Product Expiration. Then, select the notification methods and recipients as needed.

Unsubscription

Any unused quota in a compute plan automatically expires and is not refunded after the validity period ends.

If you have not used the resource plan within five days of purchase and want to unsubscribe, go to the Unsubscription Management page in the Alibaba Cloud Management Console under Expenses and Costs. On the Resource Plan tab, select the plan and complete the unsubscription.