As of MapR 5.2.1, you can create python applications for HPE Ezmeral Data Fabric Event Store using the HPE Ezmeral Data Fabric Event Store
Python client. The HPE Ezmeral Data Fabric Event Store Python client is a binding for librdkafka and the HPE Ezmeral Data Fabric Event Store C Client is
a distribution of librdkafka that works with HPE Ezmeral Data Fabric Event Store.
The HPE Ezmeral Data Fabric Event Store Python client is available in a Ecosystem Pack (MEP) starting with MEP
3.0.
The following Apache Kafka librdkafka versions are supported:
Table 1. Supported Apache Kafka librdkafka versions
| MapR release |
MEP Release |
Kafka librdkafka version |
| As of MapR 6.0.1 |
As of 5.0 |
0.11.3 |
| As of MapR 5.2.1 through 6.0.0 |
As of 3.0 |
0.9.0 |
Note: Because the HPE Ezmeral Data Fabric Event Store Python Client is dependent on the HPE Ezmeral Data Fabric Event Store C Client, the HPE Ezmeral Data Fabric Event Store C
Client must be configured before using the HPE Ezmeral Data Fabric Event Store Python Client.
When developing and running
HPE Ezmeral Data Fabric Event Store Python applications, note the following points:
- You can create producers and high-level consumers. Low-level consumers are not
supported.
- Consuming or producing topics in a Kafka cluster is not supported.
- HPE Ezmeral Data Fabric Event Store offset values start at 1, not 0.
- MapR security is supported including ACLs and ACEs for authorization. The unique Kafka
security features that are part of Apache Kafka are not supported. See Security
for more information about
MapR security features.
- User impersonation is not supported.