Nifi - Merge multiple columns to row

hi all, I am trying to import the below structure. ​ While importing, I need to merge multiple records from the same file (kind of self-join) to produce something like this. Here, all records having the same ID are being merged together into a single row. To separate the records, I have used

Re: PutHDFS with mapr

Ravi, I assume the MapR client package is working and operational and you can login to the uid running NiFi and issues the following successfuly: $ maprlogin authtest $ maprlogin print $ hdfs dfs -ls / So if those fail, fix them before you proceed. If those work, I would point the issue is

Re: NiFi Variables

Thank you for taking the time to try it! Please let me know if there's other features you would find useful. On Tue, 27 Mar 2018, 04:10 scott, wrote: > Daniel, > > That worked perfectly. Thank you for the help, and for creating this great > tool. > > Scott > > On 03/25/2018

Setting NIFI_HTTP_WEB_HOST in Docker doesn't work

Hello. I have installed NiFi Docker on a Google Cloud Debian 9 server. The server has a public IP of PUBLIC_IP (placeholder for the real IP). I ran docker pull apache/nifi and it installed the image successfully. Then, I ran: sudo docker run --name nifi -p 8080:8080 -d -e

Re: NiFi as a Web Services server

thanks Mike, you've reasoned my concerns as well. On Mon, Mar 26, 2018 at 5:57 PM, Mike Thomsen wrote: > I think you would hit two big barriers in design: > > 1. NiFi just isn't designed to be an app server for additional service > layer components a la Tomcat. > 2.

Re: how to edit queue content

If you know one of the supported scripting languages, you can probably do some of that with ExecuteScript. For example, if you wanted to drop every other flowfile in a block of 100, it'd be like this: def flowfiles = session.get(100) // Get up to 100 int index = 1 flowfiles?.each { flowFile -> if

Re: how to edit queue content

How is it going to work with e.g. 20GB of events in the queue? I'd be careful, as requirements blow up into a full db with indexes, search, and a UI on top. If one wanted to filter events, wouldn't a standard processor do the job better? Andrew On Tue, Mar 27, 2018, 12:11 AM Joe Witt

RE: Unable to create HiveConnectionPool with kerberos.

Thanks Pierre, It is working now. Mohit From: Pierre Villard Sent: 27 March 2018 13:15 To: users@nifi.apache.org Subject: Re: Unable to create HiveConnectionPool with kerberos. Hi, It needs to be the principal of the Hive server, not yours. Can you

RE: Unable to create HiveConnectionPool with kerberos.

When I try using my user prinicipal instead of hive it gives following error: SelectHiveQL[id=633d54ed-0162-1000--6fa47d56] org.apache.nifi.processors.hive.SelectHiveQL$$Lambda$523/1312347477@1c34a7fa failed to process due to java.lang.IllegalArgumentException: Kerberos principal