Github user trixpan commented on the pull request:

    https://github.com/apache/nifi/pull/180#issuecomment-173761891
  
    @mattyb149 very good point... very good point
    
    Indeed and according to a blog post on found.no(Elastic's cloud offering) 
back in January suggested the use of transport client:
    
    https://www.elastic.co/blog/found-interfacing-elasticsearch-picking-client
    
    So you are 100% right. 
    
    Seems like elastic.co has a differing view on what method to use depending 
the scope of the integration.
    
    However, to make things worse the same blog post also make statements like:
    
    > JVM update version on both the clients and the servers due to exception 
serializing 
    > and other potential minor differences between updates.
    
    Meanwhile community comments on mixed versions compatibility seem to be 
mixed bag as we 
[see](https://groups.google.com/forum/#!msg/elasticsearch/nBJ-Rz1kX6Y/flAO0zqjULcJ)
 
[here](https://github.com/elastic/elasticsearch/issues/13273#issuecomment-169189577)
    
    To be completely honest, we face the same issue with Kafka NAR, however, it 
feels like the Kafka's API policy is explicit about a strong desire to support 
older clients, while Elastic.co policy seems to be version parity. 
    
    Keen to hear what the rest of the folks say.
    
    
    *Regarding using the *Http processors:*
    
    Can the PostHTTP processor craft a JSON payload as required by the HTTP 
Bulk API or would it require people to use other processors to do it?
    
    My impression was that the pure PostHTTP cannot be used to achieve ES 
connectivity easily but I may be wrong.
    
    This is by the way the beauty of the PutElasticSearch you put together: it 
seems so simple to use! Flowfile in, elasticsearch out. **Extreme awesomeness**.
    
    From a user standpoint I guess this is what people expect to achieve when 
they request the ability to push and fetch data into/from ES. The functionality 
helps to reduce their learning curve, reduces duplication (and offloads the 
works to the development team...).



---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---

Reply via email to