[
https://issues.apache.org/jira/browse/RYA-414?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16309984#comment-16309984
]
ASF GitHub Bot commented on RYA-414:
------------------------------------
Github user jdasch commented on a diff in the pull request:
https://github.com/apache/incubator-rya/pull/256#discussion_r159482145
--- Diff:
extras/rya.geoindexing/geo.mongo/src/test/java/org/apache/rya/indexing/geotemporal/mongo/MongoITBase.java
---
@@ -1,64 +0,0 @@
-/**
--- End diff --
So it looks like you consolidated MongoITBase and MongoTestBase, which is
great. However, since MongoTestBase via EmbeddedMongoSingleton spawns new
Mongo processes, it is by definition an Integration Test and should probably be
renamed to MongoITBase for clarity. Also, tests that use it
(EntityQueryNodeTest, EventQueryNodeTest, MongoEventStorageTest, etc) should
probably be renamed with an IT suffix so they can be exercised and cleaned up
by the maven-failsafe-plugin.
> Fix inconsistent method of using MongoClient within a Sail object
> -----------------------------------------------------------------
>
> Key: RYA-414
> URL: https://issues.apache.org/jira/browse/RYA-414
> Project: Rya
> Issue Type: Bug
> Reporter: Andrew Smith
> Assignee: Kevin Chilton
> Priority: Blocker
>
> We've decided to introduce a stateful configuration object that is used to
> hold onto objects that needs to be shared across components within a Sail
> object in favor of having a static holder. The static holder limited our
> ability to use more than a single Sail object within a single JVM because all
> of those Sail objects would all have to access the same Mongo DB repository.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
