I'm a new user of elastic search and like what I've seen so far. I have
some questions on versioning support:
It seems that each update creates a new version of the document. I'm
concerned about the index grow overtime since our app will have frequent
updates on some objects.
Can I clean up the index by deleting old versions of a document?
In some cases I actually don't want to create a new version when updating
a document. Is there any API support for this in-place update?
I'm a new user of elastic search and like what I've seen so far. I have
some questions on versioning support:
It seems that each update creates a new version of the document. I'm
concerned about the index grow overtime since our app will have frequent
updates on some objects.
Can I clean up the index by deleting old versions of a document?
In some cases I actually don't want to create a new version when updating
a document. Is there any API support for this in-place update?
When you update a document, what effectively happens is that the document
gets reindexed again as if it was a new document, and the "old" document
gets marked as deleted. In the background, there is a merge process going
around that expunges deleted docs (among other things). (this is all Lucene,
btw). There isn't a way to do an "in place update", you can only "reindex" a
doc with updated values.
I'm a new user of Elasticsearch and like what I've seen so far. I have
some questions on versioning support:
It seems that each update creates a new version of the document. I'm
concerned about the index grow overtime since our app will have frequent
updates on some objects.
Can I clean up the index by deleting old versions of a document?
In some cases I actually don't want to create a new version when
updating a document. Is there any API support for this in-place update?
Thank you. This is actually the behavior that I want, most of the times. I
was just confused about the versioning feature that was introduced recently.
Can you clarify further what kind of update creates a new version and keeps
both versions that can be queried? I read the Elasticsearch Platform — Find real-time answers at scale | Elastic and I thought
that that all updates create a new version and keep the old version, instead
of marking the "old" as deleted...
On Mon, Aug 29, 2011 at 12:05 PM, Shay Banon kimchy@gmail.com wrote:
When you update a document, what effectively happens is that the document
gets reindexed again as if it was a new document, and the "old" document
gets marked as deleted. In the background, there is a merge process going
around that expunges deleted docs (among other things). (this is all Lucene,
btw). There isn't a way to do an "in place update", you can only "reindex" a
doc with updated values.
I'm a new user of Elasticsearch and like what I've seen so far. I have
some questions on versioning support:
It seems that each update creates a new version of the document. I'm
concerned about the index grow overtime since our app will have frequent
updates on some objects.
Can I clean up the index by deleting old versions of a document?
In some cases I actually don't want to create a new version when
updating a document. Is there any API support for this in-place update?
Thanks,
-Qin Zhang
Model N.
--
Qin Zhang | Senior Director, Product Development | Model N, Inc. | Direct:
650.610.4845| Fax: 650.610.4699 | qzhang@modeln.com | www.modeln.com
Versioning does not mean keeping old versions around to be searchable. It
just means that documents now have a version associated with them.
On Tue, Aug 30, 2011 at 12:02 AM, Qin Zhang qzhang@modeln.com wrote:
Thank you. This is actually the behavior that I want, most of the times.
I was just confused about the versioning feature that was introduced
recently. Can you clarify further what kind of update creates a new version
and keeps both versions that can be queried? I read the Elasticsearch Platform — Find real-time answers at scale | Elastic and I thought
that that all updates create a new version and keep the old version, instead
of marking the "old" as deleted...
On Mon, Aug 29, 2011 at 12:05 PM, Shay Banon kimchy@gmail.com wrote:
When you update a document, what effectively happens is that the document
gets reindexed again as if it was a new document, and the "old" document
gets marked as deleted. In the background, there is a merge process going
around that expunges deleted docs (among other things). (this is all Lucene,
btw). There isn't a way to do an "in place update", you can only "reindex" a
doc with updated values.
I'm a new user of Elasticsearch and like what I've seen so far. I have
some questions on versioning support:
It seems that each update creates a new version of the document. I'm
concerned about the index grow overtime since our app will have frequent
updates on some objects.
Can I clean up the index by deleting old versions of a document?
In some cases I actually don't want to create a new version when
updating a document. Is there any API support for this in-place update?
Thanks,
-Qin Zhang
Model N.
--
Qin Zhang | Senior Director, Product Development | Model N, Inc. |
Direct: 650.610.4845| Fax: 650.610.4699 | qzhang@modeln.com | www.modeln.com
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.