Member since
05-17-2016
190
Posts
46
Kudos Received
11
Solutions
My Accepted Solutions
Title | Views | Posted |
---|---|---|
1386 | 09-07-2017 06:24 PM | |
1787 | 02-24-2017 06:33 AM | |
2574 | 02-10-2017 09:18 PM | |
7058 | 01-11-2017 08:55 PM | |
4697 | 12-15-2016 06:16 PM |
10-05-2017
08:38 PM
Thanks @Karthik Narayanan, same reason that I was on a NiFi that had Avro 1.7 the logicalType doesn't work. If I remember correctly, it threw an invalid avro schema error.
... View more
10-05-2017
05:46 PM
Thanks @Karthik Narayanan, I can use them now, but the question was when we had not released NiFi 1.2
... View more
09-13-2017
02:16 PM
1 Kudo
We would not make a performance comparison of Spark and NiFi in here for this usecase since we aren't talking even running into a cluster. Since the data volume is low and processing is finite, I would prefer NiFi to be used since the entire life cycle gets reduced. + it is UI driven.
... View more
09-12-2017
06:40 PM
Understanding the problem statement here, you just have small files on the local file system and do not want to use any HDFS layer in here. Spark can, but may be an over kill for this scenario. You can start with basic java jdbc code to get this task accomplished. However, you have Apache NiFi within the HDF stack that can easily solve this problem for you. You can simply design a dataflow that will monitor for new files into your SFTP location, pull those files to NiFi as soon as they land, optionally parse them/apply some transformations, perform a mysql operation. You can use this as a reference. Additionally, NiFi can also be used as a powerful MySQL CDC options. You can see more details on this 3 part article.
... View more
09-07-2017
06:24 PM
Looks like the serde is not in the class path. Can you try add jar /usr/hdp/<version number>/hive/lib/hive-contrib-<version>.jar; and then create the table.
... View more
05-11-2017
03:26 PM
Thanks @Matt Burgess. Wanted to be sure if "replace" on the template was a dirty fix.
... View more
05-11-2017
02:53 AM
1 Kudo
Hi All, I have seen this article by @Yolanda M. Davis and this pretty much explains how to read static property file in NiFi and use these values within the flow. However, since the properties on Hive Connection pooling service doesn't support expression languages, I think we may not be able to use this approach here.
Having to migrate templates across environments, is there an approach to externalize the DHCP properties like keytab file, principal, jdbc url etc?
Thanks
-ak-
... View more
Labels:
- Labels:
-
Apache NiFi
04-18-2017
03:07 AM
Hi All,
I am trying to store CSV data to AVRO. I have issue with columns that have no value(NULL). Instead of getting nulls in the avro, it gets replaced by an empty string.
What do I need to do to make sure that empty columns in CSV are represented as nulls in my avro file.
My sample CSV looks like -
,12,"street 1"
and my avro schema for the same is
{
"type": "record",
"name": "record",
"doc": "Schema generated by Kite",
"fields": [
{
"name": "name",
"type": [
"null",
"string"
],
"default": null
},
{
"name": "age",
"type": [
"null",
"long"
],
"default": null
},
{
"name": "address",
"type": [
"null",
"string"
],
"default": null
}
]
}
To verify the conversion, if I transform my AVRO to a JSON file, I get
{
"name" : "",
"age" : 12,
"address" : "street 1"
}
But I expect it to be
{
"name" : null,
"age" : 12,
"address" : "street 1"
}
... View more
Labels:
- Labels:
-
Apache NiFi
04-04-2017
08:38 PM
@milind pandit : Can you try the same without the ; (semicolon) at the end. Not sure though.
... View more