[
https://issues.apache.org/jira/browse/NIFI-4122?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16066967#comment-16066967
]
ASF GitHub Bot commented on NIFI-4122:
--------------------------------------
Github user pvillard31 commented on a diff in the pull request:
https://github.com/apache/nifi/pull/1948#discussion_r124613250
--- Diff:
nifi-nar-bundles/nifi-mongodb-bundle/nifi-mongodb-processors/src/test/java/org/apache/nifi/processors/mongodb/GetMongoTest.java
---
@@ -200,4 +201,14 @@ public void testLimit() throws Exception {
List<MockFlowFile> flowFiles =
runner.getFlowFilesForRelationship(GetMongo.REL_SUCCESS);
flowFiles.get(0).assertContentEquals(DOCUMENTS.get(0).toJson());
}
+
+ @Test
+ public void testResultsPerFlowfile() throws Exception {
--- End diff --
This integration test assumes that we have 3 documents in the collection.
In order to have a more accurate unit test, I'd suggest to set the
``GetMongo.RESULTS_PER_FLOWFILE`` to 2 and then check that the number of flow
files sent to the SUCCESS relation ship is equal to 2 (one JSON array of two
elements, and one JSON array of one element).
> GetMongo should be able to group results into a set of flowfiles
> ----------------------------------------------------------------
>
> Key: NIFI-4122
> URL: https://issues.apache.org/jira/browse/NIFI-4122
> Project: Apache NiFi
> Issue Type: Improvement
> Reporter: Mike Thomsen
> Priority: Minor
> Labels: getmongo, mongodb, nifi
>
> GetMongo should be able to take a user-defined limit and group results by
> that size into flowfiles rather than having only the ability to do a 1:1
> relationship between result and flowfile.
> For example, if the user specifies 100, 100 results should be grouped
> together and turned into a JSON array that can be broken up later as needed.
> This need arose when doing a bulk data ingestion from Mongo. We had shy of
> 400k documents, and the 1:1 generation of flowfiles blew right through our
> limits on the content repository. Adding this feature would make it feasible
> to control that sort of behavior more thoroughly for events like bulk
> ingestion.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
