Install and Upgrade Notes
The notes in this topic contain important information about installing and upgrading Cloudera Enterprise. You should review these notes before installing or upgrading your software. For general release notes about Cloudera Enterprise, see Cloudera Enterprise 6 Release Guide.
Upgrades to Cloudera Enterprise 6.x
Using OpenJDK 11 on CDH6.3 and above requires re-installation of YARN MapReduce Framework JARs
``` 2019-07-18 14:54:52,483 ERROR [main] org.apache.hadoop.mapreduce.v2.app.MRAppMaster: Error starting MRAppMaster java.lang.NoSuchMethodError: sun.nio.ch.DirectBuffer.cleaner()Lsun/misc/Cleaner; at org.apache.hadoop.crypto.CryptoStreamUtils.freeDB(CryptoStreamUtils.java:41) at org.apache.hadoop.crypto.CryptoInputStream.freeBuffers(CryptoInputStream.java:687) at org.apache.hadoop.crypto.CryptoInputStream.close(CryptoInputStream.java:320) at java.base/java.io.FilterInputStream.close(FilterInputStream.java:180) ```
- Go to the YARN service.
- Select Actions > Install YARN MapReduce Framework JARs and click Install YARN MapReduce Framework JARs
- To verify, you will find the new MR Framework JARs under the MR Application Framework Path (default: /user/yarn/mapreduce/mr-framework/) For example:
`` hdfs dfs -ls /user/yarn/mapreduce/mr-framework/ Found 5 items -rw-r--r-- 332 yarn hadoop 215234466 2018-07-19 11:40 /user/yarn/mapreduce/mr-framework/3.0.0-cdh6.0.0-mr-framework.tar.gz -rw-r--r-- 97 yarn hadoop 263033197 2018-05-18 18:38 /user/yarn/mapreduce/mr-framework/3.0.0-cdh6.0.x-mr-framework.tar.gz -rw-r--r-- 331 yarn hadoop 222865312 2018-11-08 14:39 /user/yarn/mapreduce/mr-framework/3.0.0-cdh6.1.0-mr-framework.tar.gz -rw-r--r-- 327 yarn hadoop 232020483 2019-02-25 22:46 /user/yarn/mapreduce/mr-framework/3.0.0-cdh6.2.0-mr-framework.tar.gz -rw-r--r-- 326 yarn hadoop 234641649 2019-07-23 15:49 /user/yarn/mapreduce/mr-framework/3.0.0-cdh6.3.0-mr-framework.tar.gz ```
Cloudera Bug: CDH-81350
Restart of Impala and Hive required for Cloudera Manager 6.2 upgrade with ADLS
Cloudera Bug: OPSAPS-47436
TSB-359 Backup and Disaster Recovery (BDR) HDFS and Hive Replications will fail on clusters running Cloudera Manager 6.1.0
Backup and Disaster Recovery (BDR) HDFS and Hive Replications will fail when replicating from secured (Kerberized) source clusters to destination clusters that have been upgraded to Cloudera Manager 6.1.0.
This also affects new installations of Cloudera Manager 6.1.0 on the destination cluster if an admin restarts the Cloudera Manager service.
Products affected: Cloudera Manager Backup and Disaster Recovery in a secure (Kerberized) environment
Releases affected: Cloudera Manager 6.1.0 (when used as the destination cluster of HDFS and/or Hive replication)
Users affected: Customers using HDFS or Hive Replication
Severity (Low/Medium/High): High
Root Cause and Impact:
In HDFS and Hive Replication, Cloudera Manager first runs a process on the destination cluster to verify if the replication is possible. Due to a bug, the source cluster is treated as an insecure (non-kerberized) cluster. As a result, replication fails.
You will see the exception javax.security.sasl.SaslException: GSS initiate failed [Caused by GSSException: No valid credentials provided (Mechanism level: Fail to create credential. (63) - No service creds)] in the process stderr logs.
Immediate action required: If you use BDR, do not upgrade a destination cluster to Cloudera Manager 6.1.0. Upgrade to Cloudera Manager 6.1.1 or higher when it becomes available.
If you have already upgraded your destination cluster to Cloudera Manager to 6.1.0, use the following workaround:
- For an existing HDFS or Hive replication schedule, select .
- Save the schedule.
Please note that you will need to edit only one schedule even if you have multiple schedules.
Note: This workaround is not persistent. That is, if you restart the Cloudera Manager service, you must repeat the above workaround.
Cloudera Issue: OPSAPS-48865
Fixed in Cloudera Manager 6.1.1
Upgrades from Cloudera Enterprise 5.15 or 5.16 to 6.0x are not supported
You cannot upgrade to Cloudera Manager or CDH 6.0.0 from Cloudera Manager or CDH 5.15 or 5.16.
Upgrading to CDH 6.1.0 Enables Direct SQL mode in Hive service by default
For details about the Cloudera Manager Enable Direct SQL option, refer to Hive Metastore Database.
Upgrades from Cloudera Enterprise 6.0 Beta Release to 6.x General Release Not supported
You cannot upgrade to any Cloudera Manager or CDH 6.x general release from the Cloudera Manager or CDH 6.0 Beta release.
Cloudera Express License Enforcement
Use of Cloudera Express is limited to a total of 100 hosts running CDH6.0 or later across all environments used by an organization..
- Cloudera Manager will not allow you to add hosts to a CDH 6.x cluster if the total number of hosts across all CDH 6.x clusters will exceed 100.
- Cloudera Manager will not allow you to upgrade any cluster to CDH 6.x if the total number of managed CDH6.x cluster hosts will exceed 100. If an upgrade from Cloudera Manager 6.0 to 6.1 fails due to this limitation, you must downgrade Cloudera Manager to version 6.0, remove some hosts so that the number of hosts is less than 100, then retry the upgrade.
Affected Versions: CM 6.1 and higher
Cloudera Issue: OPSAPS-46868
Cloudera Data Science Workbench is Not Supported with Cloudera Enterprise 6.0
Cloudera Data Science Workbench is not supported with Cloudera Enterprise 6.0.x. Cloudera Data Science Workbench 1.5.0 (and higher) is supported with Cloudera Manager 6.1.x (and higher) and CDH 6.1.x (and higher).
Impala roles with SELECT or INSERT privileges receive REFRESH privileges during the upgrade
Due to the Sentry and Impala fine grained privileges feature in 5.16.0, if a role has the SELECT or INSERT privilege on an object in Impala before upgrading to CDH 5.16.0, that role will automatically get the REFRESH privilege during the upgrade.
Hue requires manual installation of psycopg2
If you are installing or upgrading to CDH 6.0.0 and using the PostgreSQL database for the Hue database, you must install psycopg2 2.5.4 or higher on all Hue hosts. SeeInstalling the psycopg2 Python package.
Cloudera Issue: OPSAPS-47080
CDH Upgrade fails to delete Solr data from HDFS
The CDH upgrade process fails to delete Solr data from HDFS and the recreated collections fail to be initialized due to the existing indexes.
- Log in to the Cloudera Manager Admin Console.
- Go to the Solr service page.
- Stop the Solr service and dependent services. Click .
- Click .
- Click .
- Start the Solr and dependent services. Click .
- Click .
Affected Versions: CDH 6.0.0
Fixed Versions: Cloudera Manager 6.0.1
Cloudera Issue: OPSAPS-47502
Package Installation of CDH Fails
When you install CDH with packages from a custom repository, ensure that the version of CDH you select for Select the version of CDH matches the version of CDH for the custom repository. Selecting the CDH version and specifying a custom repository are done during the Select Repository stage of installation.
If the versions do not match, installation fails.
Affected Versions: Cloudera Manager 6.x
Fixed Versions: N/A
Apache Issue: N/A
Cloudera Issue: OPSAPS-45703
Uninstall CDH 5 Sqoop connectors for Teradata and Netezza before upgrading to CDH 6
Sqoop includes two connectors, one for Teradata and one for Netezza. The connectors are released in separate parcels and tarballs and can be installed in Cloudera Manager or manually. The versioning of the connectors takes the form <connector_version>c<major_cdh_version>. For example, 1.6c5 refers to the connector 1.6 for CDH 5. The manifest files do not prohibit installing the CDH 5 connectors on CDH 6, but they are not compatible with CDH 6.
If you have CDH 5 connectors installed, they will not be automatically upgraded during the CDH upgrade, and they are not compatible with CDH 6, so they should be uninstalled before the upgrade. Keeping the CDH 5 connectors will not cause the upgrade to fail, but instead will cause a failure to occur during Sqoop runtime. Cloudera will release the connectors for CDH 6 at a later time.
For more information about the Teradata and Netezza connectors, go to Cloudera Enterprise Connector Documentation and choose the connector and version to see the documentation for your connector.
Unsupported Sqoop options cause upgrade failures
New fail-fast checks for unsupported options were introduced in CDH 6. Users should check the jobs stored in their Sqoop metastore and remove all unsupported options. Some unsupported options were silently ignored in earlier CDH versions during upgrades, but in CDH 6, the same options fail instantly. See the following JIRAs in Apache Sqoop Incompatible Changes:
Generated Avro code from CDH 5 should be regenerated when upgrading
Changes in logical types cause code generated in Avro with CDH 6 to differ from code generated in Avro with CDH 5. This means that old generated code will not necessarily work in CDH 6. Cloudera recommends that users regenerate their generated Avro code when upgrading.
Upgrading Apache Parquet to CDH 6
Parquet packages and the project’s group ID were renamed, and some of the class methods were removed.
If you directly consumes the Parquet API instead of using Parquet through a CDH component, your need to update and recompile your code. See Parquet API Change for details of the changes.
No HBase Replication Peer Configuration Change During Rolling Update
When doing a rolling upgrade from a CDH 6 version to a higher version, do not do any replication peer configuration changes. This includes removing a peer, adding a peer, and changing the configuration on a peer.
Oracle Database Initialization
SELECT name, value FROM v$parameter WHERE name = 'compatible'
TLS Protocol Error with OpenJDK
If you are using an older version of OpenJDK 1.8 and have enabled SSL/TLS for the Cloudera Manager Admin Console, you may encounter a TLS protocol error when connecting to the Admin Console, stating that there are no ciphers in common. This is because older versions of OpenJDK may not implement certain TLS ciphers, causing an inability to log into the Cloudera Manager Admin Console when TLS is enabled.
- Upgrade OpenJDK to a supported version of OpenJDK that is higher than version 1.8.0_181.
- If it is not possible to upgrade OpenJDK, enable less secure TLS ciphers in Cloudera Manager. You can do this by opening the /etc/default/cloudera-scm-server in a text editor and adding the following line:
export CMF_OVERRIDE_TLS_CIPHERS=<cipher_list>Where <cipher_list> is a list of TLS cipher suites separated by colons. For example:
Cloudera Bug: OPSAPS-49578
Restart Kafka after upgrading Cloudera Manager
Cloudera Bug: OPSAPS-49741
Upgrade Failure: active NameNode not found
On a Cloudera Manager cluster host running on either CentOS 7.0 or 7.1, performing a curl operation to access the NameNode web UI results in an error similar to the following if the web UI is SSL-enabled:
curl https://nn1.example.com:20102 curl: (35) Cannot communicate securely with peer: no common encryption algorithm(s).
This issue occurs on Cloudera Manager clusters running versions 6.1 and higher that exclude insecure ciphers by default. In addition, an SSL bug with curl on CentOS versions 7.0 and 7.1 prevents negotiation of a more secure SSL cipher with the HDFS NameNode web UI.
Workaround: Update yum packages to the latest available versions by running the following command:
yum update nss-util nss-sysinit nss-tools
curl --tlsv1 --ciphers rsa_aes_256_cbc_sha_256 -k -v <hostname>:<port>
- Log into the Cloudera Manager Admin Console.
- Go to the HDFS Service
- Select the Configuration tab.
- Locate the SSL/TLS Cipher Suite property
- Select Intermediate 2018 (Needed for legacy clients or Redhat 6).
- Restart the HDFS service
Cloudera Bug: CDH-81328
Upgrades to Cloudera Enterprise 5.x
Flume Kafka client incompatible changes in CDH 5.8
Due to the change of offset storage from ZooKeeper to Kafka in the CDH 5.8 Flume Kafka client, data might not be consumed by the Flume agents, or might be duplicated (if kafka.auto.offset.reset=smallest) during an upgrade to CDH 5.8.
Cloudera Issue: TSB-173
Upgrade to CDH 5.13 or higher Requires Pre-installation of Spark 2.1 or Spark 2.2
If your cluster has Spark 2.0 or Spark 2.1 installed and you want to upgrade to CDH 5.13 or higher, you must first upgrade to Spark 2.1 release 2 or later before upgrading CDH. To install these versions of Spark, do the following before running the CDH Upgrade Wizard:
- Install the Custom Service Descriptor (CSD) file. See
- Download, distribute, and activate the Parcel for the version of Spark that you are installing:
- Spark 2.1 release 2: The parcel name includes "cloudera2" in its name.
- Spark 2.2 release 1: The parcel name includes "cloudera1" in its name.
Affected versions: CDH 5.13.0 and higher
Cloudera Issue: CDH-56775
Sentry may require increased Java heap settings before upgrading CDH to 5.13
Before upgrading to CDH 5.13 or higher, you may need to increase the size of the Java heap for Sentry. A warning will be displayed during upgrade, but it its the user's responsibility to ensure this setting is adjusted properly before proceeding. See Performance Guidelines.
Affected versions: CDH 5.13 or higher
Cloudera Issue: OPSAPS-42541
Apache MapReduce Jobs May Fail During Rolling Upgrade to CDH 5.11.0 or CDH 5.11.1
2017-06-08 17:43:37,173 WARN [Socket Reader #1 for port 41187] org.apache.hadoop.ipc.Server: Unable to read call parameters for client 10.17.242.22on connection protocol org.apache.hadoop.mapred.TaskUmbilicalProtocol for rpcKind RPC_WRITABLE java.lang.ArrayIndexOutOfBoundsException: 23 at ...
This error could cause the task and the job to fail.
Avoid performing a rolling upgrade to CDH 5.11.0 or CDH 5.11.1 from CDH 5.10.x or lower. Instead, skip CDH 5.11.0 and CDH 5.11.1 if you are performing a rolling upgrade, and upgrade to CDH 5.12 or higher, or CDH 5.11.2 or higher when the release becomes available.
Cloudera Issue: DOCS-2384, TSB-241
Cloudera Manager set catalogd default jvm memory to 4G can cause out of memory error on upgrade to Cloudera Manager 5.7 or higher
After upgrading to 5.7 or higher, you might see a reduced Java heap maximum on Impala Catalog Server due to a change in its default value. Upgrading from Cloudera Manager lower than 5.7 to Cloudera Manager 5.8.2 no longer causes any effective change in the Impala Catalog Server Java Heap size.
When upgrading from Cloudera Manager 5.7 or later to Cloudera Manager 5.8.2, if the Impala Catalog Server Java Heap Size is set at the default (4GB), it is automatically changed to either 1/4 of the physical RAM on that host, or 32GB, whichever is lower. This can result in a higher or a lower heap, which could cause additional resource contention or out of memory errors, respectively.
Cloudera Issue: OPSAPS-34039