What's new in this release?
Learn about the new functionalities and improvements in Cloudera Flow Management (CFM) and how these new features benefit you.
Service Pack for CFM 2.1.4
On June 28, 2022, Cloudera released CFM 2.1.4 SP1 to fix some issues. If you previously upgraded to CFM 2.1.4.0, upgrade to CFM 2.1.4.1000-5 to access the latest version of the software. You can find the new download links in Download locations. For more information about the issues fixed in CFM 2.1.4 SP1, see Fixed issues and Fixed CVEs.
CFM 2.1.4
CFM 2.1.4 is based on Apache NiFi 1.16.0 and includes significant improvements and fixes. The most important new features and improvements of the release are:
- The flow.xml.gz file replaced with flow.json.gz
- CFM 2.1.4 NiFi started using a
flow.json.gz
instead of theflow.xml.gz
file used in all previous versions. On startup, NiFi converts the existingflow.xml.gz
to aflow.json.gz
file. If aflow.json.gz
already exists, NiFi ignores theflow.xml.gz
on startup. Changes made on the canvas will currently be written to both the newflow.json.gz
version as well as theflow.xml.gz
file; however, only theflow.json.gz
is loaded on NiFi startup. - Supported NiFi Registry databases changed
- Following some security related changes and dependency upgrades, the list of supported databases for the NiFi Registry component has been updated. Make sure you are using a supported version before upgrading to CFM 2.1.4. For more information, see Supported NiFi Registry databases.
- Kafka 0.x processors removed
- The Kafka processors previously distributed for interacting with Kafka 0.x clusters
have been removed. It means that these processors are no longer provided and supported
by Cloudera. It is recommended to switch to the processors making use of Kafka 2.x
versions.
- ConsumeKafka
- ConsumeKafka_0_10
- ConsumeKafka_0_11
- ConsumeKafkaRecord_0_10
- ConsumeKafkaRecord_0_11
- GetKafka
- PublishKafka
- PublishKafka_0_10
- PublishKafka_0_11
- PublishKafkaRecord_0_10
- PublishKafkaRecord_0_11
- PutKafka
- Some Elasticsearch processors removed
- Elasticsearch processors that were leveraging the Elasticsearch 2.0 library and the
deprecated Transport Client have been removed. It means that the below processors are no
longer provided and supported by Cloudera. It is recommended to switch to the processors
interacting with Elasticsearch over HTTP.
- FetchElasticsearch
- PutElasticsearch
- Retry at framework level
- This feature provides the ability to define a retry strategy for the flowfiles going into a given relationship for each processor.
- Clustering model and flow definition reconciliation
- A disconnected node in a NiFi cluster will no longer cause the UI to be read-only for flow changes and a reconciliation will be executed on the flow definition when the node rejoins the cluster.
- Snowflake Connection Pool controller service
- This controller service can be used in combination with the JDBC processors to easily interact with Snowflake for both pulling and pushing data.
- Query Salesforce Object processor
- A new processor is available that can be used to ingest data from Salesforce with incremental changes.
- OAuth 2 integration with InvokeHTTP processor
- It is now possible to configure an OAuth2 token provider with InvokeHTTP to easily interact with services requiring OAuth2 authentication.