Hey, i'm using NIFI 1.22 and preparing NIFI flow for upgrade to 2.0.0 version. Currently, using  PutElasticsearchHttpRecord in order to index flowfiles which contain array of nested jsons. This processor sends each flowfile as whole, does not require splits to single jsons or defining schema. This processor is deprecated and it suggested to move to PutElasticsearchRecord. My data is very much complex and variant so I tried to avoid defining schema and to use  the Inffer Schema strategy of JsonTreeReader in  PutElasticsearchRecord. But the processor is failing on type conversion of the first field in the first json of the array. It seems that it cannot handle nested jsons in array (while splitting it to single jsons worked). In addition, my implemetation need to deal the following limitation:  1. Processing a large dataset (TBs), so it need to be fast. 2. Avoiding spliting the array of jsons into a single json per flow file, as it cause content claim issues (too many large files and small files in the flow). 3. Avoiding managing schema, as the data is much variant. Love to hear how is best to Index this data to Elasticsearch. ... View more