Fixed Issues in Hue

Review the list of Hue issues that are resolved in Cloudera Runtime 7.2.18.

CDPD-59802: Hue is unable to process parallel requests

Due to an issue with Impala file reader not getting closed after use, it caused a leak in the connection pool. Because of the leak, the Hue Query Processor could not read Impala event files after a few queries. This issue has been fixed by ensuring that the file reader is closed, and ingestion of Impala query works as expected.

CDPD-45438: Hue Importer does not validate the column name; shows an error if the column name contains a space

If column names contained spaces in the file you were trying to import into Hue, then you encountered the following error: AnalysisException: Invalid column/field name. When you fixed the column name and proceeded with importing the file, the rename operation also failed. This issue has been fixed. Hue now validates the column names to verify whether they contain spaces and does not allow you to proceed without correcting the column names.

TSB 2023-703: Risk of data loss when using Hue S3 File Browser

This issue has been fixed.

CDPD-54714: SSO does not work while logging in from the Hue UI

Earlier, due to a missing configuration in Cloudera Manager, SSO did not work when you enabled Knox as an authentication backend and when Hue was in HA mode. This issue has been fixed.

CDPD-61550: ABFS File Browser lists a maximum of 5000 objects

Earlier, when you browsed ADLS Gen 2 storage using Hue’s ABFS File Browser, it displayed only 5000 objects. This issue has been fixed by including a continuation token in the response header (x-ms-continuation).

CDPD-56024: The last status of a workflow task does not reset in Hue

Earlier, when you reran an Oozie job or workflow from the Hue web interface, the task status continued to display the old status and did not reset. This has been fixed.

CDPD-59896: Hue fails to clean up documents if Oozie is on the blocked list of services

Earlier, when you ran the desktop_document_cleanup script to clean old documents, and if you added Oozie to the blocked list of apps in the Hue Advanced Configuration Snippet, the script failed. To run the script successfully, you had to remove the Oozie service from the app_blacklist property manually. This issue has been resolved. You no longer need to manually remove Oozie from the blocked list of apps.

CDPD-61589: Downloading a file from ABFS using Hue returns a corrupted file

Earlier, downloading a file from ABFS using the ABFS File Browser returned a corrupted file because of reading 1 extra byte for every 1 MB chunk size. This issue has been fixed.

CDPD-50554: Unable to deselect the checkbox on Hue's Job Browser page after a transition

Earlier, when you selected a job from the Running list from the Job Browser page and terminated it, the job remained selected after transitioning to the Completed list. This issue has been fixed.

CDPD-50493: Sample data from Table Browser in Hue launches expensive queries from the Impala Views

Earlier, when you tried to view sample data from the Table Browser in Hue, Hue automatically ran the following query: SELECT * FROM <database>.<table> LIMIT 100 from tables and views. To generate data from views, Hue fetched all rows and then displayed 100, based on the limit. This issue has been fixed. Hue no longer fetches any sample data from views.

CDPD-57894: Unable to filter Oozie tasks based on status in Hue

The ability to filter Oozie tasks based on their status from the Job Browser > Schedules page that was removed has now been added.

CDPD-61651: Schema and tables are not displayed in the Phoenix Editor as Phoenix is case-sensitive

Schema and table names in Phoenix are case-sensitive. Earlier, Hue used the UPPER() function while processing queries, resulting in issues when running Phoenix queries. This issue has been resolved by removing the UPPER() function for databases and tables.

CDPD-61606: Potential data loss while moving files or folders using the Hue S3 File Browser

When you attempted to copy or move files or folders using the Hue File Browser, you might have encountered data loss if you did not wait for the Move to modal to load completely before clicking Move, or if the source and destination paths were the same. This issue has been fixed.

CDPD-58714: Cannot rerun multiple jobs (scheduled tasks of a coordinator job)

Earlier, when you selected more than one job on the Jobs > Schedule page in Hue, the Rerun button was disabled. You could only run one job. This issue has been fixed. You can now rerun multiple jobs in Hue.

CDPD-57919: YARN does not send the active ResourceManager to Hue

If there is a ResourceManager failover, then Hue continues to send requests to the old ResourceManager instance (which is on standby mode). When you tried to terminate the job, the kill job command failed with the following error: YARN RM returned a failed response: This is standby RM. The redirect url is: /ws/v1/cluster/apps/application_1687323161493_0002/state?doAs=admin (error 307). You had to restart Hue to temporarily resolve the issue. This issue has been fixed.

CDPD-58012: Spark SQL editor (Spark3 with Livy) does not reflect output when rendering array and map columns

Earlier, the Spark SQL editor in Hue did not display column results for array, map, and struct complex data types. The same query that included these data types worked fine from Hive and Impala editors and showed results as expected for the concerned columns. This issue has been fixed.

CDPD-48059: Unable to view and access S3 buckets from Hue configured with RAZ

Earlier, when you configured access to AWS S3 buckets using RAZ in the following regions, you could not access S3 buckets from the Hue File Browser: cn-*, eu-central, ap-northeast-2, ap-south-1, us-east-2, ca-central and eu-west. Due to a bug in AWS Boto SDK, accessing S3 buckets from Hue using RAZ was blocked. This issue has been fixed.

CDPD-55430: Debug-level logging is missing in Hue server logs

The Hue server logs contained only INFO-level information and DEBUG-level information was missing. Cloudera has fixed multiple logging issues, overall. Now, only INFO-level logging is on by default. You can turn on DEBUG-level logging from Cloudera Manager by selecting the Enable Django Debug Mode option in Hue configurations.

CDPD-60510: Query Processor fails to start on a FedRAMP cluster

Earlier, the Query Processor service did not start on a FedRAMP cluster, and you may have seen the following error in the logs: ERROR io.dropwizard.cli.ServerCommand - Unable to start server, shutting down java.io.IOException: Invalid keystore format. This happened because in FIPS mode, the Query Processor service expects to find its truststore in the Java KeyStore (JKS) format, even though it might be using the Bouncy Castle library (which supports the bcfks format). This issue has been fixed by adding the keystore_type configuration (Java Keystore Type field) in Cloudera Manager through which you can specify the appropriate JKS type (in this case bcfks), which enables the FIPS mode to work.