How can Index a Filesystem?


(pumacy112) #1

In elasticsearch 1.5 are Rivers deprecated. The FSRiver Plugin is not working with ElasticSearch V. 1.6.
How is a good way to Index a File System (example: C.\temp) without the FSRiver?


How do Index a Filesystem to own Mapping
(Ed) #2

You will have to use Logstash and File input to grab your files


(pumacy112) #3

It gives a Sample for Index a File System with Logstash?


(David Pilato) #4

It won't work until we publish a Tika codec plugin IMO.


(Ed) #5

The simplest would be

input{
file {
   path=>["c:\temp\*"]
}
}
output{
stdout {
codec => "rubydebug"
}
}

https://www.elastic.co/guide/en/logstash/current/plugins-inputs-file.html


(pumacy112) #6

It is also possible to write the file datas to a own Mapping?


(pumacy112) #7

Is there a good way to do indexing a file system about the Java API from Elasticsearch?


(Ed) #8

Sure you just need to issue a "PUT/POST" request

via any language or even do it with CURL
https://www.elastic.co/guide/en/elasticsearch/reference/current/docs-index_.html


(David Pilato) #9

But what kind of files are you expecting?


(pumacy112) #10

All kind of files are possible. I will scan a File Url like c:\temp. And with Tika Parser parse all file datas and write it to my Index with own Mapping. And i will find out the best method how to do it.
Its no Documented how can be automatically


(poonam) #11

HI,
I have a requirement where I am reading file path from database and fetching files from file server.
Like KBid=1 KBfilepath="filepath" and KBstatus="ready".
Now I need to feed file and also other contents like KBID to elasticsearch.
How do I do this logtash or with river.am using ES1.5.2


(David Pilato) #12

You can't at this moment.
You need to write your own code IMO.

Or wait for LS tika codec / filter. But it's not there yet.


(Open Semantic Search) #13

You can use the files connector of Open Semantic Search: http://www.opensemanticsearch.org/etl/elasticsearch


(system) #14