In general, Talend certifies a specific release version for a given Big Data (Hadoop) Distribution vendor. These are typically what is recommended for use for that vendor. For incremental upgrades and service packs by a given vendor, Talend relies on the vendors' compatibility statements to ensure the proper running and execution of the Talend software. Where compatibility is stated, Talend also supports that version under our Support SLA. If an incompatibility should be verified by the Hadoop vendor, then Talend considers that a re-test and upgrade may be necessary.
If support for the Hadoop distribution you want to use is not yet available in your Talend Studio, it may be available via an update.
-
Dynamic distributions for HDP 3.x and CDH 6.x are in technical preview.
- On the version list of the distributions, some versions are labelled Builtin. These versions were added by Talend via the Dynamic distribution mechanism and delivered with the Studio when the Studio was released. They are certified by Talend, thus officially supported and ready to use.
Hadoop distributions
: officially supported.
: not officially supported.
: the Kerberos kinit option is supported by the Studio.
: the Kerberos kinit option and the Kerberos keytab option are both supported by the Studio.
1. This enables the connection between the Studio and a custom Hadoop distribution not yet officially supported in the Studio. For further information, see the sections describing how to connect to a custom Hadoop distribution of the Talend Big Data Getting Started Guide or the documentation of any related component that creates the connection to a Hadoop distribution, such as tHDFSConnection.
2. For this MapR version, the MapR security ticket mechanism is supported by the Studio.
3. In MapR 6, HBase support has been removed in favor of MapR-DB. For more information, see https://maprdocs.mapr.com/home/ReleaseNotes/install_upgrade_notes.html.
4. The supported Amazon EMR versions for the tAmazonEMRManage component are 4.0.0, 4.1.0, 4.2.0, 4.5.0, 4.6.0, 4.9.2 and 5.11.0.
5. Officially supported Cloudera version. Other Cloudera versions could be compatible as part of Cloudera backward compatibility on Cloudera V5.X branches.
6. HDFS in Qubole is ephemeral and thus often used to temporarily store the intermediate data before this data is transferred to a persistent storage system, such as S3. While a Talend Job for Apache Spark can write data in this HDFS system, tHDFSConfiguration does not support it.
7. This Cloud Big Data platform is not available in the Metadata wizards in the Repository but is supported in components only .
8. The Profiling perspective does not support the Embedded connection mode on Hive distributions. This mode is available mainly for test purposes done by Hadoop developers. The studio may not be able to run correctly with the embedded mode.
HBase | HCatalog | HDFS | Hive 8 | Oozie | Pig | Spark | Sqoop | Talend MapReduce | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|
Standalone | YARN | |||||||||||
HDP | V2.4.0 |
|
|
|
|
|
|
V1.6
|
|
|
||
V2.5.0 |
|
|
|
|
|
|
V1.6
|
|
|
|||
V2.6.0 |
|
|
|
|
|
|
V1.6 and V2.1
|
|
|
|||
Cloudera | CDH 5.5 (YARN mode) |
|
|
|
|
|
|
V1.5 | V1.5
|
|
|
|
CDH 5.6 (YARN mode) |
|
|
|
|
|
|
V1.5 | V1.5
|
|
|
||
CDH 5.7 (YARN mode) |
|
|
|
|
|
|
V1.6 | V1.6
|
|
|
||
CDH 5.8 (YARN mode) |
|
|
|
|
|
|
V1.6 | V1.6 and V2.0
|
|
|
||
CDH 5.10 (YARN mode) |
|
|
|
|
|
|
V1.6 and V2.1 | V1.6 and V2.1
|
|
|
||
CDH 5.12.1 (YARN mode) |
|
|
|
|
|
|
V2.2 | V2.2 |
|
|
||
CDH 5.13.0 |
|
|
|
|
|
|
V2.2 | V2.2
|
|
|
||
MapR | 5.0.0 (YARN mode) 2 |
|
|
|
|
|
|
V1.3 | V1.3
|
|
|
|
5.1.0 (YARN mode) 2 |
|
|
|
|
|
|
V1.5 | V1.5
|
|
|
||
5.2.0 (YARN mode) 2 |
|
|
|
|
|
|
V1.6 | V1.6
|
|
|
||
6.0.0 (YARN mode) 2 |
|
|
|
|
|
V2.1 | V2.1 |
|
|
|||
6.0.1 / MEP 5.0 (YARN mode) 2 |
|
|
|
|
|
V2.2 | V2.2 |
|
|
|||
Amazon EMR 4 | EMR 4.5.0 (Apache 2.7.2) |
|
V1.6 | |||||||||
EMR 4.6.0 (Apache 2.7.2) |
|
V1.6 | ||||||||||
EMR 5.0.0 (Apache 2.7.2) |
|
V2.0 |