Does anyone know if it is possible to do aggregations that are similar to
the "nested" aggregation but for parent/child relationships? I have the
following example:
parent doc:
{
"_id": "123456",
"name": "I am the parent"
}
I would like to execute a terms aggregation by language and a terms
sub-aggregation by country. Is that possible? I've done it with nested
relationships instead of parent/child, but we update this data often, and
nested relationships are harder to insert.
This is a feature that we are considering but parent/child joins are
typically much more expensive to do than nested (block) joins so we still
need to figure out if we can expose such an aggregation with a reasonable
cost.
Does anyone know if it is possible to do aggregations that are similar to
the "nested" aggregation but for parent/child relationships? I have the
following example:
parent doc:
{
"_id": "123456",
"name": "I am the parent"
}
I would like to execute a terms aggregation by language and a terms
sub-aggregation by country. Is that possible? I've done it with nested
relationships instead of parent/child, but we update this data often, and
nested relationships are harder to insert.
This is a feature that we are considering but parent/child joins are
typically much more expensive to do than nested (block) joins so we still
need to figure out if we can expose such an aggregation with a reasonable
cost.
Does anyone know if it is possible to do aggregations that are similar to
the "nested" aggregation but for parent/child relationships? I have the
following example:
parent doc:
{
"_id": "123456",
"name": "I am the parent"
}
I would like to execute a terms aggregation by language and a terms
sub-aggregation by country. Is that possible? I've done it with nested
relationships instead of parent/child, but we update this data often, and
nested relationships are harder to insert.
I'm interested in similar architecture. I have company -> employee ->
availability structure. Any update on grandchild aggregation?
While on 2nd level I get data, on 3rd level I'm always getting doc_count 0.
Or is this completely different issue?
Best,
Pete
On Tuesday, 20 May 2014 16:04:00 UTC+2, Jad wrote:
This is a feature that we are considering but parent/child joins are
typically much more expensive to do than nested (block) joins so we still
need to figure out if we can expose such an aggregation with a reasonable
cost.
On Mon, May 19, 2014 at 7:03 PM, Jad <jna...@gmail.com <javascript:>>
wrote:
Hi,
Does anyone know if it is possible to do aggregations that are similar
to the "nested" aggregation but for parent/child relationships? I have the
following example:
parent doc:
{
"_id": "123456",
"name": "I am the parent"
}
I would like to execute a terms aggregation by language and a terms
sub-aggregation by country. Is that possible? I've done it with nested
relationships instead of parent/child, but we update this data often, and
nested relationships are harder to insert.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.