ASF GitHub Bot commented on RYA-316:

Github user pujav65 commented on a diff in the pull request:

    --- Diff: 
    @@ -64,14 +68,14 @@
         public void createIndices(final DBCollection coll){
             BasicDBObject doc = new BasicDBObject();
    -        doc.put(SUBJECT, 1);
    -        doc.put(PREDICATE, 1);
    +        doc.put(SUBJECT_HASH, 1);
    +        doc.put(PREDICATE_HASH, 1);
    --- End diff --
    When the Mongo db backend was first implemented, you could only do indices 
over two fields-- the first is the primary index, the second the secondary 
index.  That may have changed since.  The indices we originally had were 
subject, predicate, object, and then subject/predicate, predicate/object, and 
object/subject.  The not including object type might be a bug, but I had 
thought that was addressed at some point.  Also one could argue that the single 
field indices were redundant-- I had wanted to test to see but never got around 
to it.
    If you can now index over more than two fields, then we might want to 
revisit this.  

> Long LineStrings break MongoDB ingest
> -------------------------------------
>                 Key: RYA-316
>                 URL: https://issues.apache.org/jira/browse/RYA-316
>             Project: Rya
>          Issue Type: Bug
>          Components: dao
>            Reporter: Aaron Mihalik
>            Assignee: Andrew Smith
> MongoDB will reject statements they contain very long linestrings.  
> Basically, the mongodb index key is limited to 1024 chars, so the insert will 
> fail if the literal is longer.
> [Here is some example 
> code|https://github.com/amihalik/rya-mongo-debugging/blob/master/src/main/java/com/github/amihalik/rya/mongo/debugging/linestring/LoadLineString.java].
>   I think the inserts will work if you use 10 points, but fail if you use 
> linestrings with 100 points.

This message was sent by Atlassian JIRA

Reply via email to