Joseph Percivall commented on MINIFI-66:

[~aldrin] I may be confused but I don't think this is the right direction. I 
would prefer we go the direction of a combination of exposing the 
agent-specific (bootstrap) properties to the config transformer[1] and either 
just preserving the properties on registry update (like was done in 
MINFI-434[2]) or add them to the registry data model.

I don't think we want to go this direction because as a user I want to be able 
to have the ability to change as much as possible within a warm config change 
as possible. With the end-goal of natively supporting having various different 
classes of agents, if I have a bug within the "instance process configuration" 
of a class of agent, I don't want to have to manually modify it on all of them. 

Ideally, I think we'd be trying to go the other way and exposing more of the 
configuration that is able to be automatically warm re-deployed as possible. 
That said, is the idea to migrate the flow portion out of the config.yml to 
another file that is dynamically updateable?

[1] https://issues.apache.org/jira/browse/MINIFI-424
[2] https://issues.apache.org/jira/browse/MINIFI-434


> Migrate non-flow properties from config.yml to bootstrap
> --------------------------------------------------------
>                 Key: MINIFI-66
>                 URL: https://issues.apache.org/jira/browse/MINIFI-66
>             Project: Apache NiFi MiNiFi
>          Issue Type: Improvement
>          Components: Agent Configuration/Installation, Processing 
> Configuration
>    Affects Versions: 0.0.1
>            Reporter: Aldrin Piri
>            Assignee: Aldrin Piri
>            Priority: Major
> To facilitate greater ease in configuring instances, it would be helpful to 
> have the config.yml be a descriptor only of processing flow with the actual 
> instance process configuration (system properties, file locations, etc) being 
> migrated to something like bootstrap.conf.

This message was sent by Atlassian JIRA

Reply via email to