[
https://issues.apache.org/jira/browse/MINIFI-171?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15822256#comment-15822256
]
ASF GitHub Bot commented on MINIFI-171:
---------------------------------------
Github user apiri commented on a diff in the pull request:
https://github.com/apache/nifi-minifi-cpp/pull/37#discussion_r96061120
--- Diff: conf/minifi.properties ---
@@ -16,6 +16,7 @@
# Core Properties #
nifi.version=0.2.0
nifi.flow.configuration.file=./conf/config.yml
+nifi.accept.dynamic.properties=true
--- End diff --
Any reason why we shouldn't support dynamic properties by default? I
understand we are trying to delineate between just blindly accepting new
properties, but feel like the model present in NiFi is the best where
processors can choose to support it.
> Dynamic Properties support for processors
> -----------------------------------------
>
> Key: MINIFI-171
> URL: https://issues.apache.org/jira/browse/MINIFI-171
> Project: Apache NiFi MiNiFi
> Issue Type: Improvement
> Components: C++
> Affects Versions: cpp-0.1.0
> Reporter: Jeremy Dyer
> Assignee: Jeremy Dyer
> Fix For: cpp-0.2.0
>
>
> Currently any Property read from the config.yml file that is not explicitly
> defined in the processor's implementation will be ignored by Processor.cpp
> when reading the configurations. This prevents any dynamic property from
> being defined in the config.yml and passed to the processor at runtime.
> Certain processors rely heavily on the concept of dynamic properties that are
> passed to them at runtime to handle things like setting dynamic properties
> based on properties that are declared. All of these possibilities cannot be
> handled upfront so there should be a mechanism, most likely in Processor.cpp,
> that allows for a list of dynamicProperties that are parsed form the
> config.yml file to be stored and accessed by the underlying processor
> implementation at runtime and use them as the processor desires.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
