Cloudera Docs

Fixed issues and improvements

Learn about the fixed issues, improvements, and changes functionality in this release of Cloudera Streaming Analytics - Kubernetes Operator.

Fixed issues

CSA-5816: Cannot export an SSB project to GitHub that has secrets
Previously, exporting projects containing secrets to GitHub resulted in an empty dialog on Force Push. This issue is resolved by improving error handling when pushing sensitive data and updating data source naming.
CSA-5858: V2 API no longer allows executing jobs in session mode
Updated the API and connectors to no longer permit session mode execution.
CSA-5923: Stop with savepoint API hangs on Kubernetes
The "Stop with savepoint" API hung on Kubernetes due to a change in savepoint handling behavior. The API now verifies the upgradeSavepointPath to ensure a new savepoint is generated before stopping the job.
CSA-5930: Cannot stop session cluster even when only finished jobs are present on Kubernetes
Session clusters with only finished jobs could not be stopped due to the FlinkSessionJobs job not being automatically cleaned up. This is fixed by checking for globally terminal state and deleting the jobs so the session cluster may be terminated.
CSA-5933: Sampling fails on Kubernetes when incoming row kind is not INSERT
Sampling failed on Kubernetes for non-INSERT row kinds. This issue is fixed by formatting all incoming rows as INSERT.

Improvements

Learn about the fixed issues and improvements in this release of Cloudera Streaming Analytics - Kubernetes Operator.

Improvements

Learn about the change in certain functionality of Apache Flink and Cloudera SQL Stream Builder that has resulted in a change in behavior from the previously released version of Cloudera Streaming Analytics - Kubernetes Operator.

CSA-5940 - Add optimization to use single source for sampling in SSB
Summary: Optimized Flink SQL jobs to use a single source for sampling, reducing resource consumption and preventing job failures caused by duplicated source tasks.
CSA-5863 - Add option to specify truststore type for Kafka connections
Summary: Added support for specifying truststore and keystore types, such as JKS, PKCS12, or BCFKS, for Kafka connections in Global Config → Sampling Kafka and Data Sources → Kafka.

Behavioral changes

Learn about the change in certain functionality of Flink and Cloudera SQL Stream Builder that has resulted in a change in behavior from the previously released version of Cloudera Streaming Analytics - Kubernetes Operator.

Behavioral changes

Summary: [CSA-5372] - Use Apache Kerby KrbConfig for more robust EncType config, refactor keytab validation

Previous behavior:

The final Kerberos encryption type that is used by Cloudera SQL Stream Builder when generating a keytab was set according to the following precedence:

  1. Values extracted from the keytab value given in the source.enctype.keytab Cloudera SQL Stream Builder configuration property.
  2. Default values defined in the Cloudera SQL Stream Builder.

New behavior:

The final Kerberos encryption type that is used by Cloudera SQL Stream Builder when generating a keytab is set according to the following precedence:

  1. Values defined in the krb5.conf file.
  2. Values extracted from the keytab value given in the source.enctype.keytab Cloudera SQL Stream Builder configuration property.
  3. Default values defined in the Cloudera SQL Stream Builder.

The following encryption types are the default in Cloudera SQL Stream Builder:

  • aes256-cts-hmac-sha1-96
  • aes128-cts-hmac-sha1-96
  • des3-cbc-sha1
  • arcfour-hmac
  • des-hmac-sha1
  • des-cbc-md5
  • des-cbc-crc