David,
Attributes are always Strings. I don’t think you can change how they are
represented in AttributesToJSON.
My recommendation would be to not use that processor and instead use
ReplaceText with a Replacement Strategy as Always. Then you can build the JSON
to look however you want by doing
You could probably use a JOLT Transform on the NiFi side (although I’m no JOLT
expert).
From an Elasticsearch point of view, if you create an index template [1] for
the destination index with mappings to store the fields as numbers, I think
Elasticsearch will do the conversion for you (pretty s
Hi
I am using the attributesToJSON processor to create a JSON flowfile which I
then send to Elastic, I have noticed that some of my attributes which ar
numbers such as file size always come out of the attributesToJSON processor as
string values (ie with double quotes around them), therefore when
>>
>> You could use ReplaceText to build a new Json document from flow file
>> attributes. Something like...
>>
>> Replacement Text = { "myField" : "${attribute1}" }
>>
>> -Bryan
>>
>> On Saturday, October 3, 2015, Sumanth Chi
cument from flow file
> attributes. Something like...
>
> Replacement Text = { "myField" : "${attribute1}" }
>
> -Bryan
>
> On Saturday, October 3, 2015, Sumanth Chinthagunta
> wrote:
>
>> do we have attributes to json processor ?
&g
Hello,
You could use ReplaceText to build a new Json document from flow file
attributes. Something like...
Replacement Text = { "myField" : "${attribute1}" }
-Bryan
On Saturday, October 3, 2015, Sumanth Chinthagunta
wrote:
> do we have attributes to json processor ?
&g
do we have attributes to json processor ?
I am thinking to use it along with ExtractText where the matching data is
stored in attributes. now I need to convert those attributes into json flowFile.
if we don’t have such processor, any ideas to use, existing processors to
compose what I needed