You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@nifi.apache.org by "Nicholas Carenza (JIRA)" <ji...@apache.org> on 2017/03/29 23:15:41 UTC

[jira] [Commented] (NIFI-3654) PutKinesisFirehose Needs A Separate Outgoing Relationship for Retryable Errors

    [ https://issues.apache.org/jira/browse/NIFI-3654?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15948069#comment-15948069 ] 

Nicholas Carenza commented on NIFI-3654:
----------------------------------------

the processor does write failure attributes that i can use

aws.kinesis.firehose.error.message
aws.kinesis.firehose.error.code

> PutKinesisFirehose Needs A Separate Outgoing Relationship for Retryable Errors
> ------------------------------------------------------------------------------
>
>                 Key: NIFI-3654
>                 URL: https://issues.apache.org/jira/browse/NIFI-3654
>             Project: Apache NiFi
>          Issue Type: Improvement
>            Reporter: Nicholas Carenza
>            Priority: Minor
>
> Just like the InvokeHttp processor understands what kinds of errors are worth retrying and which are not, so should this processor.
> For example: the error when a file is over the max size of 1000kb is unrecoverable and there is no point in retrying so output to failure. Server unavailable, output to retry... and maybe administrative yield?



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