BTW, i will suggest you to avoid using parent/child relationships if your 
index contains more than million documents per shard
besides the obvious problems with maintaining parent/child/grandchild 
relationships, parent/child queries tend to be quite slow and id_cache 
(used to store parent/child ids) will consume a lot of java heap space



On Friday, February 21, 2014 10:22:05 PM UTC, Michael D. Moffitt wrote:
>
> Hello,
>
> I'm trying to create parent/child relationships between documents that are 
> all of the same type.  This seems to be working OK, except that to enable 
> routing, I have to give "top-level" docs a bogus parent (currently ID 0):
>
>      https://gist.github.com/mmoffitt/9144694
>
> Can I somehow omit parent and have it use the child's ID instead?  More 
> importantly, are there more serious problems here, since each child's 
> parent will be routed based the grandparent's ID (and not its own)?
>

-- 
You received this message because you are subscribed to the Google Groups 
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/elasticsearch/71cbb9aa-39ff-4a7d-bdcc-bc48c03d7429%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to