elastic-query-export

Export Data from ElasticSearch to CSV by Raw or Lucene Query (e.g. from Kibana). Works with ElasticSearch 6+ (OpenSearch works too) and makes use of ElasticSearch's Scroll API and Go's concurrency possibilities to work as fast as possible.

Install

Download a pre-compiled binary for your operating system from here: https://github.com/pteich/elastic-query-export/releases You need just this binary. It works on OSX (Darwin), Linux and Windows.

Arch

yay -S elastic-query-export-bin

General usage

es-query-export -c "http://localhost:9200" -i "logstash-*" --start="2019-04-04T12:15:00" --fields="RemoteHost,RequestTime,Timestamp,RequestUri,RequestProtocol,Agent" -q "RequestUri:*export*"

CLI Options

Flag	Default
-h --help		show help
-v --version		show version
-c --connect	http://localhost:9200	URI to ElasticSearch instance
-i --index	logs-*	name of index to use, use globbing characters * to match multiple
-q --query		Lucene query to match documents (same as in Kibana)
--fields		define a comma separated list of fields to export
-o --outfile	output.csv	name of output file, you can use - as filename to output data to stdout and pipe it to other commands
-f --outformat	csv	format of the output data: possible values csv, json, raw
-r --rawquery		optional raw ElasticSearch query JSON string
-s --start		optional start date - Format: YYYY-MM-DDThh:mm:ss.SSSZ. or any other Elasticsearch default format
-e --end		optional end date - Format: YYYY-MM-DDThh:mm:ss.SSSZ. or any other Elasticsearch default format
--timefield		optional time field to use, default to @timestamp
--verifySSL	false	optional define how to handle SSL certificates
--user		optional username
--pass		optional password
--size	1000	size of the scroll window, the more the faster the export works but it adds more pressure on your nodes
--trace	false	enable trace mode to debug queries send to ElasticSearch

Output Formats

• csv - all or selected fields separated by comma (,) with field names in the first line
• json - all or selected fields as JSON objects, one per line
• raw - JSON dump of matching documents including id, index and _source field containing the document data. One document as JSON object per line.

Pipe output to other commands

Since v1.6.0 you can provide - as filename and send output to stdout. This can be used to pipe it to other commands like so:

es-query-export -start="2019-04-04T12:15:00" -q "RequestUri:*export*" -outfile - | aws s3 cp - s3://mybucket/stream.csv