# Amazon EMR Serverless release versions
An Amazon EMR release is a set of open source applications from the big data ecosystem. Each release includes big data applications, components, and features that you select to have Amazon EMR Serverless deploy and configure when you run your job.
With Amazon EMR 6.6.0 and higher, deploy EMR Serverless. This deployment option isn't available with earlier Amazon EMR release versions. When you submit your job, specify one of the following supported releases.
**Topics**
+ [`AWS runtime for Apache Spark` (emr-spark-8.0-preview)](release-version-emr-spark-8.0-preview.md)
+ [EMR Serverless 7.12.0](release-version-7120.md)
+ [EMR Serverless 7.11.0](release-version-7110.md)
+ [EMR Serverless 7.10.0](release-version-7100.md)
+ [EMR Serverless 7.9.0](release-version-790.md)
+ [EMR Serverless 7.8.0](release-version-780.md)
+ [EMR Serverless 7.7.0](release-version-770.md)
+ [EMR Serverless 7.6.0](release-version-760.md)
+ [EMR Serverless 7.5.0](release-version-750.md)
+ [EMR Serverless 7.4.0](release-version-740.md)
+ [EMR Serverless 7.3.0](release-version-730.md)
+ [EMR Serverless 7.2.0](release-version-720.md)
+ [EMR Serverless 7.1.0](release-version-710.md)
+ [EMR Serverless 7.0.0](release-version-700.md)
+ [EMR Serverless 6.15.0](release-version-6150.md)
+ [EMR Serverless 6.14.0](release-version-6140.md)
+ [EMR Serverless 6.13.0](release-version-6130.md)
+ [EMR Serverless 6.12.0](release-version-6120.md)
+ [EMR Serverless 6.11.0](release-version-6110.md)
+ [EMR Serverless 6.10.0](release-version-6100.md)
+ [EMR Serverless 6.9.0](release-version-690.md)
+ [EMR Serverless 6.8.0](release-version-680.md)
+ [EMR Serverless 6.7.0](release-version-670.md)
+ [EMR Serverless 6.6.0](release-version-660.md)
# `AWS runtime for Apache Spark` (emr-spark-8.0-preview)
The following table lists the application versions available with `AWS runtime for Apache Spark` (emr-spark-8.0-preview).
**Application version information**
| Application | Version |
| --- | --- |
| Spark | 4.0.1-amzn-0 |
****`AWS runtime for Apache Spark` (emr-spark-8.0-preview) release notes****
+ **Preview release** – This is a preview release of `AWS runtime for Apache Spark` featuring Apache Spark 4.0.1. This preview is available on EMR Serverless only.
+ **Regional Availability** - This preview release is available in all AWS Regions where EMR Serverless is available, except China and AWS GovCloud (US) regions.
+ **Application version information** - This release ships with the following application versions:
+ AWS SDK for Java 2.35.5, 1.12.792
+ Python **3.9**, 3.11, 3.12
+ Scala 2.13.16
+ AmazonCloudWatchAgent 1.300034.0-amzn-0
+ Delta 4.0.0-amzn-0-spark
+ Iceberg 1.10.0-amzn-spark-0
+ This release ships with Amazon Corretto **17** (built on OpenJDK) by default for applications that support Corretto 17 (JDK 17).
+ **Preview limitations** - The following capabilities are not available in this preview release:
+ **Interactive and Integration Features**: SageMaker Unified Studio, EMR Studio integration, Spark Connect, Livy, and JupyterEnterpriseGateway are not supported.
+ **Table Formats and Access Control**: Hudi, Delta Universal Format, and fine-grained access control (FGAC) with row-level or column-level filtering and DDL/DML operators are not supported.
+ **Data Connectors**: spark-sql-kinesis, emr-dynamodb, and spark-redshift connectors are not available.
+ **History Server**: The Persistent Spark History Server is not available in this preview release. Users can still access the live Spark UI to monitor and debug active serverless jobs in real-time.
+ **Specialized Features**: Materialized Views are not available.
+ **Preview capabilities** - You can test the following capabilities in this preview release. This preview release is not recommended for production workloads:
+ **SQL Features**: ANSI SQL mode with stricter type handling, SQL PIPE syntax (\$1>) for chaining operations, VARIANT data type for semi-structured JSON data, SQL scripting with control flow statements and session variables, and SQL user-defined functions.
+ **Streaming Enhancements**: Arbitrary Stateful Processing API v2 with transformWithState operator, State Data Source Reader for queryable streaming state (experimental), and enhanced state store with improved RocksDB changelog checkpointing.
+ **Table Format Support**: Apache Iceberg v3 with VARIANT data type support, AWS S3 Tables integration, and Full Table Access (FTA) with AWS Lake Formation for Iceberg, Delta Lake, and Hive tables.
+ **Additional Documentation** - For additional Apache Spark documentation, see [Apache Spark 4.0.1 Release Documentation](https://spark.apache.org/releases/spark-release-4-0-1.html).
To get started with Apache Spark 4.0.1 preview, create an EMR Serverless application using the AWS CLI:
```
aws emr-serverless create-application --type spark \
--release-label emr-spark-8.0-preview \
--region us-east-1 --name spark4-preview
```
# EMR Serverless 7.12.0
The following table lists the application versions available with EMR Serverless 7.12.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.6 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 7.12.0 release notes**
+ **New features**
+ **Serverless storage for EMR Serverless** – Amazon EMR serverless introduces serverless storage, with EMR release 7.12 and later, that eliminates local disk provisioning for Apache Spark workloads. EMR Serverless automatically handles intermediate data operation such as shuffle with no storage charges. Serverless storage decouples storage from compute, allowing Spark to release workers immediately when idle rather than keeping them active to preserve temporary data. To learn more, see [Serverless storage](https://docs.aws.amazon.com/emr/latest/EMR-Serverless-UserGuide/jobs-serverless-storage.html).
+ **Iceberg Materialized Views** - Starting Amazon EMR 7.12.0, Amazon EMR Spark supports creation and management of Iceberg Materialized Views (MV)
+ **Hudi Full Table Access** - Starting Amazon EMR 7.12.0, Amazon EMR now supports Full Table Access (FTA) control for Apache Hudi in Apache Spark based on your policies defined in Lake Formation. This feature enables read and write operations from your Amazon EMR Spark jobs on Lake Formation registered tables when the job role has full table access.
+ **Iceberg version upgrade** - Amazon EMR 7.12.0 supports Apache Iceberg version 1.10
# EMR Serverless 7.11.0
The following table lists the application versions available with EMR Serverless 7.11.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.6 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 7.11.0 release notes**
+ **Maximum Job execution time** – The maximum value for `executionTimeoutMinutes` in `StartJobRun` action for BATCH jobs is 7 days from this release onwards. `executionTimeoutMinutes` can no longer be set to `0` i.e. no timeout, for batch job runs.
# EMR Serverless 7.10.0
The following table lists the application versions available with EMR Serverless 7.10.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.5 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 7.10.0 release notes**
+ **Metrics for EMR Serverless** – Monitoring metrics are restructured to focus on the `ApplicationName` and `JobName` dimensions. Older metrics will no longer be updated, going forward. For more information, refer to [Monitoring EMR Serverless applications and jobs](app-job-metrics.html).
# EMR Serverless 7.9.0
The following table lists the application versions available with EMR Serverless 7.9.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.5 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 7.8.0
The following table lists the application versions available with EMR Serverless 7.8.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.4 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 7.7.0
The following table lists the application versions available with EMR Serverless 7.7.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.3 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 7.6.0
The following table lists the application versions available with EMR Serverless 7.6.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.3 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 7.5.0
The following table lists the application versions available with EMR Serverless 7.5.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.2 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 7.4.0
The following table lists the application versions available with EMR Serverless 7.4.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.2 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 7.3.0
The following table lists the application versions available with EMR Serverless 7.3.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.1 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 7.3.0 release notes**
+ **Job concurrency and queuing with EMR Serverless** – Job concurrency and queuing is enabled by default when you create a new EMR Serverless application on Amazon EMR release 7.3.0 or higher. For more information, refer to [Job concurrency and queuing for an EMR Serverless application](applications-concurrency-queuing.md), which details how to get started with concurrency and queuing and also contains a list of feature considerations.
# EMR Serverless 7.2.0
The following table lists the application versions available with EMR Serverless 7.2.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.1 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 7.2.0 release notes**
+ **Lake Formation with EMR Serverless** – you can now use AWS Lake Formation to apply fine-grained access controls on Data Catalog tables that are backed by S3. This capability lets you configure table, row, column, and cell level access controls for read queries within your EMR Serverless Spark jobs. For more information, refer to [Using EMR Serverless with AWS Lake Formation for fine-grained access control](emr-serverless-lf-enable.md) and [Considerations and limitations](emr-serverless-lf-enable-considerations.md).
# EMR Serverless 7.1.0
The following table lists the application versions available with EMR Serverless 7.1.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.0 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 7.0.0
The following table lists the application versions available with EMR Serverless 7.0.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.5.0 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 6.15.0
The following table lists the application versions available with EMR Serverless 6.15.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.4.1 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 6.15.0 release notes**
+ **TLS support** – With Amazon EMR Serverless releases 6.15.0 and higher, enable mutual-TLS encrypted communication between workers in your Spark job runs. When enabled, EMR Serverless automatically generates a unique certificate for each worker that it provisions under a job runs that workers utilize during TLS handshake to authenticate each other and establish an encrypted channel to process data securely. For more information about mutual-TLS encryption, refer to [Inter-worker encryption](https://docs.aws.amazon.com/emr/latest/EMR-Serverless-UserGuide/interworker-encryption.html).
# EMR Serverless 6.14.0
The following table lists the application versions available with EMR Serverless 6.14.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.4.1 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 6.13.0
The following table lists the application versions available with EMR Serverless 6.13.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.4.1 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 6.12.0
The following table lists the application versions available with EMR Serverless 6.12.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.4.0 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
# EMR Serverless 6.11.0
The following table lists the application versions available with EMR Serverless 6.11.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.3.2 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 6.11.0 release notes**
+ **[Access S3 resources in other accounts](jobs-s3-access.md#jobs-s3-access-how-to-assumed-role-multiple)** - With releases 6.11.0 and higher, you can configure multiple IAM roles to assume when you access Amazon S3 buckets in different AWS accounts from EMR Serverless.
# EMR Serverless 6.10.0
The following table lists the application versions available with EMR Serverless 6.10.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.3.1 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 6.10.0 release notes**
+ For EMR Serverless applications with release 6.10.0 or higher, the default value for the `spark.dynamicAllocation.maxExecutors` property is `infinity`. Earlier releases default to `100`. For more information, refer to [Spark job properties](jobs-spark.md#spark-defaults).
# EMR Serverless 6.9.0
The following table lists the application versions available with EMR Serverless 6.9.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.3.0 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.10.2 |
**EMR Serverless 6.9.0 release notes**
+ The Amazon Redshift integration for Apache Spark is included in Amazon EMR releases 6.9.0 and later. Previously an open-source tool, the native integration is a Spark connector that you can use to build Apache Spark applications that read from and write to data in Amazon Redshift and Amazon Redshift Serverless. For more information, see [Using Amazon Redshift integration for Apache Spark on Amazon EMR Serverless](emr-spark-redshift.md).
+ EMR Serverless release 6.9.0 adds support for AWS Graviton2 (arm64) architecture. You can use the `architecture` parameter for the `create-application` and `update-application` APIs to choose the arm64 architecture. For more information, refer to [Amazon EMR Serverless architecture options](architecture.md).
+ You can now export, import, query, and join Amazon DynamoDB tables directly from your EMR Serverless Spark and Hive applications. For more information, refer to [Connecting to DynamoDB with Amazon EMR Serverless](using-ddb-connector.md).
**Known issues**
+ If you use the Amazon Redshift integration for Apache Spark and have a time, timetz, timestamp, or timestamptz with microsecond precision in Parquet format, the connector rounds the time values to the nearest millisecond value. As a workaround, use the text unload format `unload_s3_format` parameter.
# EMR Serverless 6.8.0
The following table lists the application versions available with EMR Serverless 6.8.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.3.0 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.9.2 |
# EMR Serverless 6.7.0
The following table lists the application versions available with EMR Serverless 6.7.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.2.1 |
| Apache Hive | 3.1.3 |
| Apache Tez | 0.9.2 |
## Engine-specific changes, enhancements, and resolved issues
The following table lists a new engine-specific feature.
| Change | Description |
| --- | --- |
| Feature | Tez scheduler now supports preemption of Tez task instead of preemption of container |
# EMR Serverless 6.6.0
The following table lists the application versions available with EMR Serverless 6.6.0.
| Application | Version |
| --- | --- |
| Apache Spark | 3.2.0 |
| Apache Hive | 3.1.2 |
| Apache Tez | 0.9.2 |
**EMR Serverless initial release notes**
+ EMR Serverless supports the Spark configuration classification `spark-defaults`. This classification changes values in Spark's `spark-defaults.conf` XML file. Configuration classifications allow you to customize applications. For more information, refer to [Configure applications](https://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-configure-apps.html).
+ EMR Serverless supports the Hive configuration classifications `hive-site`, `tez-site`, `emrfs-site`, and `core-site`. This classification can change the values in Hive's `hive-site.xml` file, Tez's `tez-site.xml` file, Amazon EMR's EMRFS settings, or Hadoop's `core-site.xml` file, respectively. Configuration classifications allow you to customize applications. For more information, refer to [Configure applications](https://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-configure-apps.html).
**Engine-specific changes, enhancements, and resolved issues**
+ The following table lists Hive and Tez backports.
**Hive and Tez changes**
[\[See the AWS documentation website for more details\]](http://docs.aws.amazon.com/emr/latest/EMR-Serverless-UserGuide/release-version-660.html)