[ 
https://issues.apache.org/jira/browse/NIFI-3682?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15981436#comment-15981436
 ] 

ASF GitHub Bot commented on NIFI-3682:
--------------------------------------

Github user joewitt commented on the issue:

    https://github.com/apache/nifi/pull/1682
  
    @markap14 ok have reviewed it quite a bit and much happier with the results.
    
    1. 'Schema Write Strategy' default should be schema.name rather than 
schema.text SchemaRegistryService:73 is where this is set
    2. QueryRecord with a GrokReader where schema access is 'Use String fields 
from Grok expression' what do you do for the Writer?  It cannot lookup the 
schema by name unless you've registered it and if you've done that then the 
grok expression inferred schema is a bit odd. 
    3. Hortonworks Schema Registry access strategy should provide a link to the 
hortonworks schema registry in github so others can know the serde approach for 
prepending header bytes to cover protocol version, schema identifier, and 
schema version.  This is here: https://github.com/hortonworks/registry
    



> Add "Schema Access Strategy" to Record Readers and Writers
> ----------------------------------------------------------
>
>                 Key: NIFI-3682
>                 URL: https://issues.apache.org/jira/browse/NIFI-3682
>             Project: Apache NiFi
>          Issue Type: Improvement
>          Components: Extensions
>            Reporter: Mark Payne
>            Assignee: Mark Payne
>             Fix For: 1.2.0
>
>
> Currently the record readers are mostly configured with a Schema Registry 
> service and the name of the schema. We should instead allow user to choose 
> one of several strategies for determining the schema: Schema Registry + 
> schema.name attribute, Schema Registry + identifier and version embedded at 
> start of record/stream, avro.schema attribute, embedded schema for cases like 
> Avro where the schema can be embedded in the content itself.
> On the writer side, we should also expose these options in order to convey 
> the schema information to others.



--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

Reply via email to