Known issues in 7.1.9 CHF 4

You must be aware of the known issues and limitations, the areas of impact, and workaround in Cloudera Runtime 7.1.9 CHF 4.

After upgrading the cluster from CDP Private Cloud Base 7.1.8 to CDP Private Cloud Base 7.1.9 and Ozone is in the Non-HA environment, an exception message is observed during the finalization of the Ozone upgrade.
During the finalization of the upgrade, ClassNotFoundException for org.cloudera.log4j.redactor.RedactorAppender class was identified. The error message is harmless as the upgrade is successful. The error existed previously and does not affect the Ozone service and its operation.
None.
When using S3A committer fs.s3a.committer.name=directory with fs.s3a.committer.staging.conflict-mode=replace to write to FSO buckets, the client fails with the following error.
DIRECTORY_NOT_FOUND org.apache.hadoop.ozone.om.exceptions.OMException: Failed to find parent directory of xxxxxxxx at org.apache.hadoop.ozone.om.request.file.OMFileRequest.getParentID(OMFileRequest.java:1008) at org.apache.hadoop.ozone.om.request.file.OMFileRequest.getParentID(OMFileRequest.java:958) at org.apache.hadoop.ozone.om.request.file.OMFileRequest.getParentId(OMFileRequest.java:1038) at org.apache.hadoop.ozone.om.request.s3.multipart.S3MultipartUploadCompleteRequestWithFSO.getDBOzoneKey(S3MultipartUploadCompleteRequestWithFSO.java:114) at org.apache.hadoop.ozone.om.request.s3.multipart.S3MultipartUploadCompleteRequest.validateAndUpdateCache(S3MultipartUploadCompleteRequest.java:157) at org.apache.hadoop.ozone.protocolPB.OzoneManagerRequestHandler.handleWriteRequest(OzoneManagerRequestHandler.java:378) at org.apache.hadoop.ozone.om.ratis.OzoneManagerStateMachine.runCommand(OzoneManagerStateMachine.java:568) at org.apache.hadoop.ozone.om.ratis.OzoneManagerStateMachine.lambda$1(OzoneManagerStateMachine.java:363) at java.base/java.util.concurrent.CompletableFuture$AsyncSupply.run(CompletableFuture.java:1700) at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128) at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628) at java.base/java.lang.Thread.run(Thread.java:834)
This occurs because S3A uses multipart upload to commit job results in a batch. The staging committer's replace mode deletes the target directory before completing MPU. The problem is that FSO does not create intermediate directories during MPU, it does only for regular file/dir/key requests.
Use fs.s3a.committer.name=magic for ** affected versions.
OPSAPS-69846: If Ozone is installed with custom kerberos principals for its roles, operations on encrypted buckets can fail as Ranger KMS does not have its proxy users and groups configured for the custom S3 Gateway user.
Add the following configurations in Ranger-kms safety valve based on the custom s3g user. In this case , the user is s3gfoo0. The parameters are hadoop.kms.proxyuser.s3gfoo0.hosts = * hadoop.kms.proxyuser.s3gfoo0.groups = *
CDPD-66508: Shallow listing is enabled by default in 7.1.9. There is a bug in shallow listing that causes the below error when listing an empty directory in a LEGACY/OBS bucket:

Error when listing an empty directory in a LEGACY/OBS bucket: mkdir: getFileStatus on s3a://testbucket/data/test: com.amazonaws.services.s3.model.AmazonS3Exception: Server Error (Service: Amazon S3; Status Code: 500; Error Code: 500 Server Error; Request ID: null; S3 Extended Request ID: null; Proxy: null), S3 Extended Request ID: null:500 Server Error: Server Error (Service: Amazon S3; Status Code: 500; Error Code: 500 Server Error; Request ID: null; S3 Extended Request ID: null; Proxy: null)

In S3 gateway log: Caused by: java.lang.IndexOutOfBoundsException: Index 0 out of bounds for length 0 at java.base/jdk.internal.util.Preconditions.outOfBounds(Preconditions.java:64) at java.base/jdk.internal.util.Preconditions.outOfBoundsCheckIndex(Preconditions.java:70) at java.base/jdk.internal.util.Preconditions.checkIndex(Preconditions.java:248) at java.base/java.util.Objects.checkIndex(Objects.java:372) at java.base/java.util.ArrayList.remove(ArrayList.java:535) at org.apache.hadoop.ozone.client.OzoneBucket$KeyIterator.getNextShallowListOfKeys(OzoneBucket.java:1234) at org.apache.hadoop.ozone.client.OzoneBucket$KeyIterator.getNextListOfKeys(OzoneBucket.java:1136) at org.apache.hadoop.ozone.client.OzoneBucket$KeyIterator.hasNext(OzoneBucket.java:1110) at org.apache.hadoop.ozone.s3.endpoint.BucketEndpoint.get(BucketEndpoint.java:208) at jdk.internal.reflect.GeneratedMethodAccessor90.invoke(Unknown Source)

Disable shallow listing by performing the following steps:
  1. Log in to Cloudera Manager
  2. Navigate to Clusters
  3. Select the Ozone service
  4. Go to Configurations
  5. In S3 Gateway Advanced Configuration Snippet (Safety Valve) for ozone-conf/ozone-site.xml, set ozone.s3g.list-keys.shallow.enabled = false.
CDPD-64394: OzoneManager may fail to start and have logs java.lang.IllegalArgumentException: Trying to set updateID to XXX which is not greater than the current value of XXX for OMKeyInfo{XXX} in OzoneManager log file.
Currently, this issue cannot be auto recovered. You must contact Cloudera support.
CDPD-65801: This is an intermittent issue in native RocksDB tool which causes corruption to in-memory RocksDB metadata.
Set ozone.om.snapshot.load.native.lib to false and restart the OM.
CDPD-66142: When Solr is slow/down, Solr takes lot of time to respond to Recon Heatmap query or sometimes doesn't respond at all which makes Recon heatmap trying to load the heatmap data forever. This issue will be taken up in future releases and solution could be to introduce a health check for Solr or timeout the Recon query to Solr and show a meaningful message over Recon UI -> "Solr is not responding"
  1. Stop Recon
  2. Restart Solr
  3. Start Recon
CDPD-66247: TestOzoneFileSystem.testListStatusOnKeyNameContainDelimiter is intermittent
None
CDPD-66382: When Bucket layout is LEGACY and ozone.om.enable.filesystem.paths property is set to true, then delete will not work completely if keyName contains "/".
None.
CDPD-66252: du space calculation support for OBS and LEGACY (fsPath disabled).
du space for OBS buckets and LEGACY(fspath disabled) can be seen using CLI command.
OPSAPS-69539: CDP Runtime 7.1.9 does not support Oracle JDK 8u401 or OpenJDK 1.8.0_402 (8u402). Some services will fail to start. This can be a problem on RHEL 9.x as version 8u402 is the default OpenJDK 8 installed by the OS.
Workaround is to install an earlier version of JDK 8. For example Oracle jdk-8u291 / 1.8.0_291, or OpenJDK 8u292 / 1.8.0_292.
A fresh install of 7.1.9 CHF 2 does not allow user to bypass the Setup Database screen for YARN Queue Manager
YARN Queue Manager in Cloudera Data Platform (CDP) Private Cloud Base 7.1.9 CHF 2 does not require you to install a PostGres database, therefore users should be able to skip the Setup Database screen. With this known issue, users who are conducting a fresh install of 7.1.9 CHF 2 are not able to bypass the Setup Database screen as expected.
  1. When conducting a fresh install of YARN Queue Manager in 7.1.9 CHF 2, you must ensure that you have both CDP and Cloudera Manager upgraded to 7.1.9 CHF 2.
  2. When you reach the Setup Database screen in the Cloudera Manager installation wizard for Queue Manager, enter any dummy values for the following fields:
    1. Database name: configstore
    2. Database Username: dbuser
    3. Database Password: dbpassword
    YARN Queue Manager will not connect to PostGres with the above details and will fall back to the embedded database.
  3. Run the following script command in a browser console to enable the Continue button:

    document.querySelector('.btn.next').removeAttribute('disabled');

  4. Click Continue and proceed with the YARN Queue Manager installation.
  5. Restart YARN Queue Manager.
CDPD-61524: Ozone Storage Container Manager fails to start on upgrading from CDP Private Cloud Base 7.1.6 to 7.1.9 CHF1. Also, if you have upgraded from CDP Private Cloud Base 7.1.6 to 7.1.7 or 7.1.8 and then to 7.1.9, the upgrade fails.
None. Cloudera recommends you to reach out to the Support before performing the upgrade to CDP Private Cloud Base 7.1.9.
CDPD-62254: Ozone is not supported on SLES15 with CHF1.
If your cluster has Ozone, Cloudera recommends you to not upgrade to 7.1.9 CHF1.
CDPD-62834: Status of the deleted table is seen as ACTIVE in Atlas after the completion of navigator2atlas migration process
The status of the deleted table displays as ACTIVE.
None
CDPD-62837: During the navigator2atlas process, the hive_storagedesc is incomplete in Atlas
For the hive_storagedesc entity, some of the attributes are not getting populated.
None
CDPD-63690: RuntimeException encountered when generating snapshotDiff report between 2 snapshots
When snapshot feature is enabled, KeyDeletingService, SSTFilteringService and SnapDiff thread fall into a deadlock when accessing Snapshot Cache.
Restart the Ozone Manager.
CDPD-64238: Snapshot diff request failing when setting ozone.om.snapshot.db.max.open.files=-1
When snapshot feature is enabled, KeyDeletingService, SSTFilteringService and SnapDiff thread fall into a deadlock when accessing Snapshot Cache.
Restart the Ozone Manager.

Technical Service Bulletins

TSB (2024-749): Possible Ozone Snapshot Chain Corruption in 7.1.9.0
The Apache Ozone (Ozone) snapshots feature provides the ability to save consistent and immutable copies of the Ozone namespace. A chain of snapshots is tracked and maintained in the Ozone RocksDB that is utilized to calculate the difference between two snapshots required for the snapDiff API. A race condition on the snapshot delete path was found during testing. This condition might result in two consecutive snapshots pointing to the same parent, which breaks the snapshot chain.
Upstream JIRA
HDDS-10524, HDDS-10590, and HDDS-9198
Knowledge article
For the latest update on this issue see the corresponding Knowledge article: TSB 2024-749: Possible Ozone Snapshot Chain Corruption in 7.1.9.0