E-MapReduce:Scope and methods of technical support

Supported services

The EMR technical support team fully supports the EMR console — including consultation, stability issues, and product defects. Console support covers the console itself, not the services deployed inside EMR clusters. Supported console operations include:

• Purchasing an EMR cluster

• Creating an EMR cluster

• Scaling an EMR cluster

• Adding services to an EMR cluster

• Configuring auto scaling rules

• Using the monitoring and alerting feature

• Developing data

The support team also fully supports EMR's self-developed services, including consultation, stability issues, and product defects:

• SmartData

• Bigboot

• ESS

• DLF-Auth

Assisted services

For the following service categories, the EMR technical support team assists in troubleshooting but does not hold full ownership of the outcome.

Open-source services in EMR

The support team assists in identifying and troubleshooting issues with open-source services. Solutions are provided on an iterative schedule aligned with the product plan. Assisted open-source services include:

• Hadoop services: Hadoop Distributed File System (HDFS), YARN, MapReduce

• Data integration: Flume, Sqoop

• Data storage: HDFS, HBase, Kudu, Iceberg, Delta Lake, Hudi, Kafka

• Compute engines: Spark, Hive, Tez, Druid, Flink, Phoenix

• Online analytical processing (OLAP): Presto, StarRocks, Doris, ClickHouse, Impala

• Other: ZooKeeper, Knox, Ganglia, OpenLDAP, Hue, Zeppelin, Livy, Superset, Ranger, Alluxio, Oozie

Alibaba Cloud services that EMR depends on

The support team assists in troubleshooting issues that originate in underlying Alibaba Cloud services. Defects in those services are handled by the respective service teams. Assisted services include:

• Computing: Elastic Compute Service (ECS), Container Service for Kubernetes (ACK) — for example, insufficient inventory or kernel defects

• Storage: Elastic Block Storage (EBS), Object Storage Service (OSS) — for example, storage, read, or write exceptions or data inconsistency

• Network: Virtual Private Cloud (VPC) — for example, misconfigured security groups or network fluctuations

• Observability: CloudMonitor, Simple Log Service — for example, alert push exceptions or log read timeouts

Services not supported

EMR does not support the following:

• Cluster and service O&M operations — EMR provides O&M features such as service management, node management, and cluster monitoring, and guarantees the feature reliability, but the support team does not perform operations on your behalf, such as capacity planning, service performance optimization, service restarts, or configuration changes.

• Issues caused by improper operations — This includes invalid custom scripts, disk space exhaustion, deletion of system directories, and improper resource use.

• Business application development — The support team does not provide consultation on or troubleshoot issues related to building applications on top of EMR, including business implementation, code development, job performance tuning, and business migration.

• Single-job troubleshooting with no cluster-level exceptions — If no obvious cluster service exceptions or service defects exist, individual job issues are outside the support scope.

• Non-standard capabilities — This includes cluster version upgrades, service configuration changes, core node scale-in, and disk capacity reduction.

• Third-party component issues — Third-party components are not covered.

• Cluster instability or unavailability caused by unexpected operations — For a list of unexpected operations, see Limits.