Cloudera Machine Learning and Data Lake compatibility
Cloudera Machine Learning maintains compatibility with the Data Lake through the Hadoop CLI addon.
This table shows the CML release and the compatible Data Lake version.
|2.0.41-b225||7.2.15 or higher|
|2.0.40-b157||7.2.14 or higher|
|2.0.40-b150||7.2.14 or higher|
|2.0.38-b126||7.2.14 or higher|
|2.0.38-b125||7.2.14 or higher|
|2.0.38-b121||7.2.14 or higher|
|2.0.38-b101||7.2.14 or higher|
|2.0.36-b121||7.2.14 or higher|
|2.0.36-b118||7.2.14 or higher|
|2.0.34-b116||7.2.14 or higher|
|2.0.32-b123||7.2.14 or higher|
|2.0.30-b114||7.2.14 or higher|
|2.0.29-b61||7.2.11 or higher|
|2.0.27-b64||7.2.11 or higher|
|2.0.26-b180||7.2.11 or higher|
|2.0.25-b110||7.2.11 or higher|
|2.0.24-b100||7.2.11 or higher|
Upgraded CML deployments keep multiple Hadoop CLI addon versions that administrators can configure to maintain compatibility between CML and the Data Lake.
CML packages the Spark versions that are included in Cloudera Data Engineering (CDE), and makes them available as Runtime Addons. In CML, users can select the Spark version they want to use for each workload. The compatibility between Spark and the Data Lake is shown in Cloudera Data Engineering and Data Lake compatibility.