Unsupported Features in CDH 6.0.1
Running Apache Accumulo on top of a CDH 6.0.x cluster is not currently supported. If you try to upgrade to CDH 6.0.x you will be asked to remove the Accumulo service from your cluster. Running Accumulo on top of CDH 6 will be supported in a future release.
Cloudera Data Science Workbench
Cloudera Data Science Workbench is not supported with CDH 6.0.x. If you try to upgrade to CDH 6.0.x, you will be asked to remove the CDSW service from your cluster.
Cloudera Data Science Workbench 1.5.0 (and higher) is supported with CDH 6.1.x (and higher).
Apache Hadoop Unsupported Features
HDFS Unsupported Features
The following HDFS features are not supported in CDH 6.0.x:
- ACLs for the NFS gateway
- Aliyun Cloud Connector
- Erasure Coding
- HDFS NameNode Federation
- HDFS Router Based Federation
- HDFS truncate
- More than two NameNodes
- Openstack Swift
- Quota support for Storage Types
- SFTP FileSystem
- Upgrade Domain
- Variable length block
- ZStandard Compression Codec
YARN Unsupported Features
The following YARN features are not supported in CDH 6.0.x:
- Application Timeline Server v2 (ATSv2)
- Cgroup Memory Enforcement
- Container Resizing
- Distributed or Centralized Allocation of Opportunistic Containers
- Distributed Scheduling
- Docker on YARN (DockerContainerExecutor)
- Native Services
- New Aggregated Log File Format
- Node Labels
- Pluggable Scheduler Configuration
- Reservation REST APIs
- Resource Estimator Service
- Resource Profiles
- Rolling Log Aggregation
- (non-Zookeeper) ResourceManager State Store
- Shared Cache
- YARN Federation
- YARN WebUI v2
Apache HBase Unsupported Features
- Master hosting meta
- Cloudera does not provide support for user-provided custom coprocessors of any kind.
- Server-side encryption of HFiles. You should configure HDFS client-side encryption.
- In-memory compaction
- Visibility labels
- Stripe compaction
- Clients setting priority on operations
- Specifying a custom asynchronous connection implementation
Apache Hive Unsupported Features
The following Hive features are not supported in CDH 6.0.x:
- AccumuloStorageHandler (HIVE-7068)
- ACID (HIVE-5317)
- Built-in version() function is not supported (CDH-40979)
- Cost-based Optimizer (CBO) and gathering column statistics required by CBO
- Explicit Table Locking
- HCatalog - HBase plugin
- Hive Authorization (Instead, use Apache Sentry.)
- Hive on Apache Tez
Hive Local Mode Execution
- Hive Metastore - Derby
- Hive Web Interface (HWI)
- HiveServer1 / JDBC 1
- HiveServer2 Dynamic Service Discovery (HS2 HA) (HIVE-8376)
- HiveServer2 - HTTP Mode (Use THRIFT mode.)
- HPL/SQL (HIVE-11055)
- LLAP (Live Long and Process framework)
- Scalable Dynamic Partitioning and Bucketing Optimization (HIVE-6455)
- Session-level Temporary Tables (HIVE-7090)
- Table Replication Across HCatalog Instances (HIVE-7341)
- TRUNCATE TABLE on external tables (causes Error: org.apache.spark.sql.AnalysisException)
Apache Kafka Unsupported Features
The following Kafka feature is not supported in CDH 6.0.x:
- CDK Powered by Apache Kafka supports Java based clients only. Clients developed with C, C++, Python, .NET and other languages are currently not supported.
- Idempotent and transactional capabilities in the producer are currently an unsupported beta feature given their maturity and complexity. This feature will be supported in a future release.
- Kafka Connect is included in CDH 6.0.0, but is not supported. Flume and Sqoop are proven solutions for batch and real time data loading that complement Kafka's message broker capability. See Flafka: Apache Flume Meets Apache Kafka for Event Processing for more information.
- Kafka Streams is included in CDH 6.0.0, but is not supported. Instead, use Spark and Spark Streaming have a fully functional ETL/stream processing pipeline. See Using Kafka with Apache Spark Streaming for Stream Processing for more information.
- The Kafka default authorizer is included in CDH 6.0.0, but is not supported. This includes setting ACLs and all related APIs, broker functionality, and command-line tools.
- Using Kafka with a JBOD setup is an unsupported beta feature given its maturity and complexity. Using JBOD in production will be supported only in a later release.
- The legacy Scala clients (producer and consumer) that are under the kafka.producer.* and kafka.consumer.* package are deprecated in CDH 6.0.0. See Deprecated Scala-based Client API and New Java Client API.
Apache Oozie Unsupported Features
The following Oozie feature is not supported in CDH 6.0.x:
- Conditional coordinator input logic.
Cloudera does not support Derby database to use with Oozie. You can use it for testing or debugging purposes, but Cloudera does not recommend using it in production environments.
Apache Pig Unsupported Features
Cloudera Search Unsupported Features
The following Search features are not supported in CDH 6.0.x:
- Solr SQL/JDBC
- Graph Traversal
- Cross Data Center Replication (CDCR)
- SolrCloud Autoscaling
- HDFS Federation
- Saving search results
- Solr contrib modules (Morphlines, Spark Crunch indexer, MapReduce and Lily HBase indexers are part of the Cloudera Search product itself, therefore they are supported)
- Logging Slow Queries
Apache Sentry Unsupported Features
The following Sentry features are not supported in CDH 6.0.x:
- Import and export of Sentry metadata to and from Sentry servers
- Sentry shell command line for Hive
- Relative URI paths (Known Issue)
- Object types Server and URI in
show grant role <role name> on object <object name>(Known Issue)
- ALTER and DROP privileges for Hive and Impala
In addition, as of CDH 6.0.x, Sentry policy files have been removed. See the Sentry Incompatible Changes for more information.
Apache Spark Unsupported Features
The following Spark features are not supported in CDH 6.0.x:
- Apache Spark experimental features/APIs are not supported unless stated otherwise.
- Using the JDBC Datasource API to access Hive or Impala is not supported
- ADLS not Supported for All Spark Components. Microsoft Azure Data Lake Store (ADLS) is a cloud-based filesystem that you can access through Spark applications. Spark with Kudu is not currently supported for ADLS data. (Hive on Spark is available for ADLS in CDH 5.12 and higher.)
- IPython / Jupyter notebooks is not supported. The IPython notebook system (renamed to Jupyter as of IPython 4.0) is not supported.
- Certain Spark Streaming features not supported. The mapWithState method is unsupported because it is a nascent unstable API.
- Thrift JDBC/ODBC server is not supported
- Spark SQL CLI is not supported
- GraphX is not supported
- SparkR is not supported
- Structured Streaming is not supported
- Spark cost-based optimizer (CBO) not supported
- Dynamic partition overwrite mode (spark.sql.sources.partitionOverwriteMode=dynamic) is not supported
- Running Spark on a host that is not managed by Cloudera Manager is not supported
Apache Sqoop Unsupported Features
The following Sqoop feature is not supported in CDH 6.0.x: