[ANN] Filesystem River for Elasticsearch 0.0.1

Elastic Stack Elasticsearch
1

Hey All,

I have just released the very first version of a new river : Filesystem
river.

This river helps to scan and index files from your local dir. Index is done
through the excellent mapper-attachment plugin.

It's the first release of this plugin, so use it with caution !

If you want to use it in a multi node context, you will have to share a
network dir with the same mount point name on each server.

Main site is here : http://www.pilato.fr/fsriver/
http://www.pilato.fr/fsriver/

Github here : https://github.com/dadoonet/fsriver

It's not pushed at this time in OSS Sonatype maven repo as a jar, but you
can use it as a plugin.

See README on github :
https://github.com/dadoonet/fsriver/blob/master/README.textile
https://github.com/dadoonet/fsriver/blob/master/README.textile

BTW, for this first version, I embedded the mapper attachment in the
fsriver. It could be a bad idea.

Is there any way when we define a plugin to check that another plugin (aka
dependency) is already loaded in ES ?

If so, I could remove the mapper attachment and only check when FSRiver
starts that the mapper plugin is installed.

Comments, Issues, Pull requests are welcome as usual.

Hope this could help

David.

2

I've installed the plugin. I'm trying to index my json documents directly
from the filesystem using this river. The file metadata is uploaded to
elasticsearch, but it doesn't upload the content, or it uploads it with an
unreadable encoding. How do I use the fsriver to upload json document
content?

--

3

Did you follow the latest instructions?

WARNING: starting from 0.0.3, you need to have the Attachment Plugin. It's not included anymore in the distribution.

--
David :wink:
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

Le 16 nov. 2012 à 11:14, Amy amyblarney@gmail.com a écrit :

I've installed the plugin. I'm trying to index my json documents directly from the filesystem using this river. The file metadata is uploaded to elasticsearch, but it doesn't upload the content, or it uploads it with an unreadable encoding. How do I use the fsriver to upload json document content?

--

4

Yes, I installed 0.0.2, but then just in case I installed the attachment
plugin. The files content are added to the index as a string and encoded as
base64.

On Friday, November 16, 2012 11:44:11 AM UTC, David Pilato wrote:

Did you follow the latest instructions?

WARNING: starting from 0.0.3, you need to have the Attachment Pluginhttps://github.com/elasticsearch/elasticsearch-mapper-attachments.
It's not included anymore in the distribution.

--
David :wink:
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

Le 16 nov. 2012 à 11:14, Amy <amybl...@gmail.com <javascript:>> a écrit :

I've installed the plugin. I'm trying to index my json documents directly
from the filesystem using this river. The file metadata is uploaded to
elasticsearch, but it doesn't upload the content, or it uploads it with an
unreadable encoding. How do I use the fsriver to upload json document
content?

--

--

5

Could you send me more details about your setup? Are you running on Windows, Linux, Mac? Is there any special encoding?
How did you create your index, your river ?
Can you send your query?

Can you send also one document (direct mail me) ? Il will try to index it on my side.

Thanks

--
David :wink:
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

Le 16 nov. 2012 à 13:03, Amy amyblarney@gmail.com a écrit :

Yes, I installed 0.0.2, but then just in case I installed the attachment plugin. The files content are added to the index as a string and encoded as base64.

On Friday, November 16, 2012 11:44:11 AM UTC, David Pilato wrote:
Did you follow the latest instructions?

WARNING: starting from 0.0.3, you need to have the Attachment Plugin. It's not included anymore in the distribution.

--
David :wink:
Twitter : @dadoonet / @elasticsearchfr / @scrutmydocs

Le 16 nov. 2012 à 11:14, Amy amybl...@gmail.com a écrit :

I've installed the plugin. I'm trying to index my json documents directly from the filesystem using this river. The file metadata is uploaded to elasticsearch, but it doesn't upload the content, or it uploads it with an unreadable encoding. How do I use the fsriver to upload json document content?

--

--