Cloudera Manager is installed using RPM packages. The link below shows you the location of the repositories for RHEL 7 and 8. https://docs.cloudera.com/cdp-private-cloud-base/latest/manager-release-notes/topics/chf1-cm-773.html ... View more

What behaviour are you seeing? Does the RM crash on startup or does it happen when it's been running for some time? The logs suggest that the RM cannot connect to ZooKeeper and is transitioning from active into standby state. ... View more

Hey, it looks like you're running Hadoop on Windows! I'm guessing that this is not a Cloudera package and that you're attempting to run some other flavour of Hadoop on a Windows host. The java.lang.UnsatisfiedLinkError suggests you're missing the hadoop.dll library from your path. ... View more

Unfortunately that isn't enough information to understand why the query is failing. Are you able to share the logs for the failed query? There ought to be some more information in there. ... View more

Hi Kamal, You're replying to a ticket from over a year ago. I would suggest creating a new ticket and providing some more details on what exactly you're seeing. Did you follow the suggestions in the previous replies and if so what was the outcome?   Cheers, Jim ... View more

Hi, What do the logs say on the Oozie and YARN servers show in response to this error? You may find a stack trace or more descriptive error message in the logs compared to what you see back in the browser. Try using your browser debugger to see what request is being sent to the web server. Check to see if you are seeing the expected WWW-Authenticate and Authorization headers from the server and client respectively. The top of this page gives and indication of what you should be seeing - https://docs.oracle.com/javase/10/security/part-vi-http-spnego-authentication.htm What changed on the DC exactly? Was it a patch or was the version of Windows upgraded? Cheers, Jim ... View more

Hi, The example query you have shown would work id you add a JOIN statement. This would allow you to select all the records where either of the incremental value columns has changed. For example:   select a.column_1, a.column_2, a.column_3, b.column_1, b.column_2, b.column_3 from table_a a  join table_b b on a.id = b.id where a.id > 2 or b.id > 1   The trouble with this is you're not really tracking changes and you risk getting duplicate records unless both table_a and table_b are updated at the same time. If they are updated simultaneously then it ought to be sufficient to track the changes in only one table.   If you really want to track changes to tables in a relational database you might want to consider using the CDC processor. There was a series of tutorial posts previously published on how to do this.   https://community.cloudera.com/t5/Community-Articles/Change-Data-Capture-CDC-with-Apache-NiFi-Part-1-of-3/ta-p/246623   Cheers, Jim ... View more

Hi Ravi, You could try Stackable (https://stackable.tech/) if you are looking for services and support just for Apache NiFi. Cloudera includes NiFi in CDP but as far as I am aware they don't offer support for NiFi outside of a CDP subscription.   Cheers, Jim ... View more

Agreed, I did a quick test and null values in records return null when the lookup is done. As André says, querying column_value = null does not match a null value in the column as an IS NULL statement would. Ideally you should not have a key column which contains nulls or duplicate rows since this is going to unpredictable results. ... View more

The CSV reader has a property name Null String that allows you to set the value that will appear in a CSV field to represent null.   Try setting this when you read your CSV records. I've not tested whether this passes null to a SQL query, but using a literal value like "null" in your CSV data will be interpreted as a string.  ... View more