Scroll

edit

While a search request returns a single “page” of results, the scroll API can be used to retrieve large numbers of results (or even all results) from a single search request, in much the same way as you would use a cursor on a traditional database.

Scrolling is not intended for real time user requests, but rather for processing large amounts of data, e.g. in order to reindex the contents of one index into a new index with a different configuration.

The results that are returned from a scroll request reflect the state of the index at the time that the initial search request was made, like a snapshot in time. Subsequent changes to documents (index, update or delete) will only affect later search requests.

In order to use scrolling, the initial search request should specify the scroll parameter in the query string, which tells Elasticsearch how long it should keep the “search context” alive (see Keeping the search context alive), eg ?scroll=1m.

curl -XGET 'localhost:9200/twitter/tweet/_search?scroll=1m' -d '
{
    "query": {
        "match" : {
            "title" : "elasticsearch"
        }
    }
}
'

The result from the above request includes a _scroll_id, which should be passed to the scroll API in order to retrieve the next batch of results.

curl -XGET 'localhost:9200/_search/scroll?scroll=1m' \ 
     -d    'c2Nhbjs2OzM0NDg1ODpzRlBLc0FXNlNyNm5JWUc1'  

GET or POST can be used and the URL should not include the index or type name — this is specified in the original search request instead. The scroll parameter tells Elasticsearch to keep the search context open for another 1m.

The returned _scroll_id attribute can be passed in the request body or in the query string as ?scroll_id=....

Each call to the scroll API returns the next batch of results until there are no more results left to return, ie the hits array is empty.

The initial search request and each subsequent scroll request returns a new _scroll_id — only the most recent _scroll_id should be used.

If the request specifies aggregations, only the initial search response will contain the aggregations results.

Efficient scrolling with Scroll-Scan

edit

Deep pagination with from and size — e.g. ?size=10&from=10000 — is very inefficient as (in this example) 100,000 sorted results have to be retrieved from each shard and resorted in order to return just 10 results. This process has to be repeated for every page requested.

The scroll API keeps track of which results have already been returned and so is able to return sorted results more efficiently than with deep pagination. However, sorting results (which happens by default) still has a cost.

Normally, you just want to retrieve all results and the order doesn’t matter. Scrolling can be combined with the scan search type to disable any scoring or sorting and to return results in the most efficient way possible. All that is needed is to add search_type=scan to the query string of the initial search request:

curl 'localhost:9200/twitter/tweet/_search?scroll=1m&search_type=scan' -d ' 
{
    "query": {
        "match" : {
            "title" : "elasticsearch"
        }
    }
}
'

Setting search_type to scan disables sorting and makes scrolling very efficient.

A scanning scroll request differs from a standard scroll request in four ways:

  • No score is calculated and sorting is disabled. Results are returned in the order they appear in the index.
  • Aggregations are not supported.
  • The response of the initial search request will not contain any results in the hits array. The first results will be returned by the first scroll request.
  • The size parameter controls the number of results per shard, not per request, so a size of 10 which hits 5 shards will return a maximum of 50 results per scroll request.

If you want the scoring to happen, even without sorting on it, set the track_scores parameter to true.

Keeping the search context alive

edit

The scroll parameter (passed to the search request and to every scroll request) tells Elasticsearch how long it should keep the search context alive. Its value (e.g. 1m, see Time units) does not need to be long enough to process all data — it just needs to be long enough to process the previous batch of results. Each scroll request (with the scroll parameter) sets a new expiry time.

Normally, the background merge process optimizes the index by merging together smaller segments to create new bigger segments, at which time the smaller segments are deleted. This process continues during scrolling, but an open search context prevents the old segments from being deleted while they are still in use. This is how Elasticsearch is able to return the results of the initial search request, regardless of subsequent changes to documents.

Keeping older segments alive means that more file handles are needed. Ensure that you have configured your nodes to have ample free file handles. See File Descriptors.

You can check how many search contexts are open with the nodes stats API:

curl -XGET localhost:9200/_nodes/stats/indices/search?pretty

Clear scroll API

edit

Search contexts are removed automatically either when all results have been retrieved or when the scroll timeout has been exceeded. However, you can clear a search context manually with the clear-scroll API:

curl -XDELETE localhost:9200/_search/scroll \
     -d 'c2Nhbjs2OzM0NDg1ODpzRlBLc0FXNlNyNm5JWUc1' 

The scroll_id can be passed in the request body or in the query string.

Multiple scroll IDs can be passed as comma separated values:

curl -XDELETE localhost:9200/_search/scroll \
     -d 'c2Nhbjs2OzM0NDg1ODpzRlBLc0FXNlNyNm5JWUc1,aGVuRmV0Y2g7NTsxOnkxaDZ' 

All search contexts can be cleared with the _all parameter:

curl -XDELETE localhost:9200/_search/scroll/_all