This topic describes the runtime lifecycle policy (RLP) of E-MapReduce (EMR). The policy contains rules that are configured at the granularities of product features and release versions. You can understand the policy rules in detail and make the required preparations in advance based on descriptions in this topic.
Background information
Product updates and phase-outs are common in the basic technical software field. With the rapid development of the open-source big data community, the features and architectures that are provided by new EMR versions become more diversified, powerful, and user-friendly. Various business forms, such as engine reconstruction, billing and metering, and cloud-native deployment, are emerging. As a result, the features that are provided by existing versions and service types may fail to meet increasing business requirements. The service performance in terms of application stability, resource scalability, processing performance, access security, and API compatibility may also be affected as time passes. Therefore, new EMR versions are essential to provide you with high-quality services.
Product lifecycle management helps introduce new service features and new versions based on business requirements. This helps you better keep up with the pace of technological development, enhances the efficiency of cloud computing, and allows cloud computing vendors to focus on services and prevent ineffective maintenance workloads. To provide you with enterprise-level service capabilities and enhance your insight into the product evolution, Alibaba Cloud EMR builds a lifecycle management system based on extensive external user research and internal production and research analysis. The lifecycle management system defines an RLP at the granularities of service features and release versions to help you understand policy rules in detail and make the required preparations in advance.
You are expected to participate in the product and version lifecycle management and provide feedback during service consultation, to sales representatives, or over the DingTalk service group. This helps you use cloud services in an efficient and stable manner, and manage your real-time business and cloud service development.
Notes
This topic is updated on June 11, 2025 to revise the mapping between existing released versions and their lifecycle milestones, add new announcements about certain E-MapReduce cluster versions approaching End of Service and Support (EOS), and update announcements regarding certain E-MapReduce cluster versions nearing End of Marketing (EOM).
This topic is updated on March 31, 2025 to add more information about the EOM of EMR versions and RLP of EMR V3.39.X, EMR V 5.5.X, and specific later versions.
This topic is updated on November 08, 2024 to add more information about new purchases and renewals during the EOM phase.
This topic is released on October 17, 2023.
RLP
For information about EMR version planning and EMR version definition, see Release versions.
An RLP is used during the process from the official delivery of an EMR version to users, to EOM, and finally to EOS.
The period of time for each phase of an RLP is not fixed. Before the EMR team determines to implement a phase of an RLP, the team considers various factors, such as the maturity and compatibility of a new version and the demands for version updates. If the team determines to implement a phase, the team notifies users at least three months in advance. For more information, see Related services and notifications. The following figure shows the milestones in an RLP of EMR.
Milestone | Description | Implementation time | Description | Suggestion |
General Availability (GA) | Official delivery | On the T day | A new version can be officially released and delivered for use in the production environment. Alibaba Cloud provides technical support and a service-level agreement (SLA) guarantee. For more information, see the following topics: | None. |
EOM | Version sales discontinuation | Three years after the GA date in principle | New purchases are not allowed. Alibaba Cloud provides technical support and a SLA guarantee only for existing clusters. For more information, see the following topics: | Upgrade existing clusters to later versions or migrate data of the existing clusters to EMR Serverless Spark or EMR Serverless StarRocks. For more information, see What is EMR Serverless StarRocks? and What is EMR Serverless Spark? |
EOS | End of service and support | At least one year after the EOM date | The desired EMR version reaches end of full support. Alibaba Cloud no longer provides technical support and service for existing EMR clusters and no longer provides security and stability SLA guarantees. If you do not migrate data at the earliest opportunity, you assume the security and stability risks that may be caused by existing EMR clusters. | The Alibaba Cloud EMR team no longer provides after-sales support for EMR versions that reach EOS. To prevent your business from being affected, upgrade your cluster before its version reaches EOS. |
Mappings between product versions and implementation time of each phase in the lifecycle
The following table describes the mappings between product versions and implementation time of each phase in the lifecycle. You can upgrade your cluster based on the implementation time of each phase in the lifecycle.
Note: The implementation time of each phase in the lifecycle may be adjusted due to product planning, but they will not be moved up.
Product version | GA | EOM | EOS |
1.X.X series | 2015 to 2017 | 2024-09-30 | 2025-10-20 |
2.X.X series | 2016 to 2018 | 2024-09-30 | 2025-10-20 |
3.0.X to 3.35.X | Earlier than 2021-04-21 | 2024-09-30 | 2025-10-20 |
4.0.X to 4.9.X | Earlier than 2021-04-21 | 2024-09-30 | 2025-10-20 |
5.0.X to 5.1.X | Earlier than 2021-04-21 | 2024-09-30 | 2025-10-20 |
3.36.X | 2021-07-21 | 2024-09-30 | 2025-10-20 |
5.2.X | 2021-07-16 | 2024-09-30 | 2025-10-20 |
3.37.X | 2021-08-21 | 2024-09-30 | 2025-10-20 |
5.3.X | 2021-08-21 | 2024-09-30 | 2025-10-20 |
3.38.X | 2021-10-21 | 2024-10-20 | 2025-10-20 |
5.4.X | 2021-10-21 | 2024-10-20 | 2025-10-20 |
3.39.X | 2022-02-15 | 2025-10-20 | 2026-10-20 |
5.5.X | 2022-02-15 | 2025-10-20 | 2026-10-20 |
4.10.X | 2022-03-23 | 2025-10-20 | 2026-10-20 |
3.40.X | 2022-04-21 | 2025-10-20 | 2026-10-20 |
5.6.X | 2022-04-21 | 2025-10-20 | 2026-10-20 |
3.41.X | 2022-05-15 | 2025-10-20 | 2026-10-20 |
5.7.X | 2022-05-15 | 2025-10-20 | 2026-10-20 |
3.42.X | 2022-08-05 | 2025-10-20 | 2026-10-20 |
5.8.X | 2022-08-05 | 2025-10-20 | 2026-10-20 |
3.43.X | 2022-10-14 | 2026-03-01 | 2027-03-01 |
5.9.X | 2022-10-04 | 2026-03-01 | 2027-03-01 |
3.44.X | 2022-12-01 | 2026-03-01 | 2027-03-01 |
5.10.X | 2022-12-01 | 2026-03-01 | 2027-03-01 |
3.45.X | 2023-2-28 | 2026-03-01 | 2027-03-01 |
5.11.X | 2023-2-28 | 2026-03-01 | 2027-03-01 |
3.46.X | 2023-6-1 | To be determined | To be determined |
5.12.X | 2023-6-1 | To be determined | To be determined |
3.47.X | 2023-8-3 | To be determined | To be determined |
5.13.X | 2023-8-3 | To be determined | To be determined |
3.48.X | 2023-8-17 | To be determined | To be determined |
5.14.X | 2023-8-17 | To be determined | To be determined |
3.49.X | 2023-10-27 | To be determined | To be determined |
5.15.X | 2023-10-27 | To be determined | To be determined |
3.50.X | 2024-2-19 | To be determined | To be determined |
5.16.X | 2024-2-19 | To be determined | To be determined |
3.51.X | 2024-4-23 | To be determined | To be determined |
5.17.X | 2024-4-23 | To be determined | To be determined |
3.52.X | 2024-12-4 | To be determined | To be determined |
5.18.X | 2024-12-4 | To be determined | To be determined |
Related services and notifications
Based on the preceding lifecycle policy rules, Alibaba Cloud performs the following operations:
Plans the lifecycle of each EMR version and provides related information for users during service consultation and the use of service.
Notifies users at least three months before the specific time of each key milestone by using a notification channel, such as announcements, internal messages, emails, or text messages.
Helps users evaluate the risks of job running and provides solutions before the service type or version is terminated.