It seems that HiveOnSpark is not supported in Cloudera Manager.
https://www.cloudera.com/documentation/spark2/latest/topics/spark2_known_issues.html#ki_hive_on_spark
Although I noticed someone saying that Hive version 2.2.0 does support Spark 2.x.
Can we really use Hive on Spark? (link below)
Cloudera Hive on Spark 2.x?
If you want Spark2 then it's currently not supported. Based on the CDH6.0.0 packaging info, it's using Hive 2.1.1 which is only compatible with Spark 1.6. If you are fine with using Spark 1.6 then it is. See Running Apache Hive on Spark in CDH for more information.
The version compatability can be found at Hive on Spark: Getting Started.
Related
Does flyway community version support Oracle 19c? From this URL: https://flywaydb.org/documentation/database/oracle, it is up to 19.3. My oracle database is 19.7. I am not very familiar with oracle database versions.
Any version support oracle 19.7 at all?
Thanks!
Assuming no breaking changes have appeared in Oracle 19.7, then yes, we just have not formally tested it yet. If you try, you will get a warning to this effect in the log but Flyway will function as normal. I'll update the tests and remove this warning for the next patch version.
Does anyone know if the latest Snowflake ODBC driver for Linux (v2.19.16) is compatible with RHEL 6.8?
Snowflake customers have had success using the Snowflake ODBC driver on RHEL. I would say the best thing to do is to test. As per the documentation, what is currently officially supported is CentOS 6 or higher. CentOS is very close to being RHEL without the branding and support. In particular, the library versions are the same, so binaries that work on one will work on the other.
https://docs.snowflake.net/manuals/user-guide/odbc.html#odbc-driver
I am not able to locate any docs to upgrade unmanaged cluster (running 5.16) via command line like the one below for 5.15
https://www.cloudera.com/documentation/enterprise/5-15-x/topics/cdh_ig_upgrade_command_line.html
Can someone point me to the corresponding doc for 6.1 ?
This has been answered in cloudera forum. Cross-posting to prevent link rot
You can upgrade the packages based cluster but it's very tedious
process so not recommended.
https://docs.cloudera.com/documentation/enterprise/upgrade/topics/ug_overview.html
The best way is to migrate from packages to parcels and then
upgrade.
https://docs.cloudera.com/documentation/enterprise/6/latest/topics/cm_ig_migrating_packages_to_parce...
Once you migrated you can use this page to get the CM server running
with Web UI.
https://docs.cloudera.com/documentation/enterprise/6/6.3/topics/installation.html
Regarding your ask "Is there any way by which an Unmanaged 5.x
cluster can move to CDH 6.x (may be adding it in Cloudera Manager
and then moving to 6.x)?"
I am thinking that you can build a 1 node CM server and then add these
present cluster by adding node in that CM. (Though I haven't tried
this on packages)
I have Artifactory 5.4 running and would like to upgrade it to the latest available one: 6.3.3. Do I need to upgrade it to some interim version or can I go straight to 6.3.3 ?
You should be able to upgrade from version 5 to 6 without the need for an interim version.
Before doing so, however, it is strongly suggested that you
Do a complete system export. If at any time you decide to roll back to the older version, you can use the export to reproduce the system in its entirety.
Backup your database.
I have a use case where I have revolution R on my desktop and want to invoke/talk to spark and spark sql deployed in a hadoop cluster(also have sparkR installed there).Any suggestions on how to proceed on this.I heard it can be done if spark is in stand alone mode.But I want with sparn in yarn mode.
You can checkout the sparkR package. This also helps in connecting to the Spark SQL as you had required
https://spark.apache.org/docs/1.5.1/sparkr.html