Since Elasticsearch rivers are deprecated this is a simple node process that will follow a couchdb changes feed and updates elasticsearch.
You can also include an optional mapper.
npm install couch2elastic4sync -g
rc is used to set to variables. For example, create a .couch2elastic4sync file with the following
database=http://localhost:5984/idx-edm-v5
elasticsearch=http://elastic-1.com:9200/idx-edm-v5/listing
or pass the config file path explicity
couch2elastic4sync --config=path/to/configfile
Alternatively, the config file can be in JSON
{
"database": "http://localhost:5984/idx-edm-v5",
"elasticsearch": "http://elastic-1.com:9200/idx-edm-v5/listing"
}
You can also destructure some of the urls to help configuration management
couch=http://localhost:5984
database=idx-edm-v5
elasticsearch=http://elastic-1.com:9200
indexName=idx-edm-v5
indexType=listing
To load all the documents into elasticsearch, run
couch2elastic4sync load
To keep a sync process going, run
couch2elastic4sync
couch2elastic4sync id 19404098
where it is the id from couch
Sometimes there are docs in elasticsearch, that dont match any docs in couch. This task cleans those up (removes them) from elasticsearch.
couch2elastic4sync cleanup
A mapper
function can be passed from the config to format documents before being put to ElasticSearch:
database=http://localhost:5984/idx-edm-v5
elasticsearch=http://elastic-1.com:9200/idx-edm-v5/listing
mapper=path/to/my-mapper.js
Where my-mapper.js could be something like
module.exports = function (doc) {
// apply formatting here
return doc
}
If the function returns empty, the document is filtered-out
MIT