OpenSearch/docs/reference/search/run-a-search.asciidoc

292 lines
6.4 KiB
Plaintext

[[run-a-search]]
== Run a search
You can use the <<search-search,search API>> to search data stored in
one or more {es} indices.
The API can runs two types of searches, depending on how you provide
<<search-query,queries>>:
<<run-uri-search,URI searches>>::
Queries are provided through a query parameter. URI searches tend to be
simpler and best suited for testing.
<<run-request-body-search,Request body searches>>::
Queries are provided through the JSON body of the API request. These queries
are written in <<query-dsl,Query DSL>>. We recommend using request body
searches in most production use cases.
[WARNING]
====
If you specify a query in both the URI and request body, the search API request
runs only the URI query.
====
[discrete]
[[run-uri-search]]
=== Run a URI search
You can use the search API's <<search-api-query-params-q,`q` query string
parameter>> to run a search in the request's URI. The `q` parameter only accepts
queries written in Lucene's <<query-string-syntax,query string syntax>>.
.*Example*
[%collapsible]
====
To get started, ingest or add some data to an {es} index.
The following <<docs-bulk,bulk API>> request adds some example user log data to
the `user_logs_000001` index.
[source,console]
----
PUT /user_logs_000001/_bulk?refresh
{"index":{"_index" : "user_logs_000001", "_id" : "1"}}
{ "@timestamp": "2020-12-06T11:04:05.000Z", "user": { "id": "vlb44hny" }, "message": "Login attempt failed" }
{"index":{"_index" : "user_logs_000001", "_id" : "2"}}
{ "@timestamp": "2020-12-07T11:06:07.000Z", "user": { "id": "8a4f500d" }, "message": "Login successful" }
{"index":{"_index" : "user_logs_000001", "_id" : "3"}}
{ "@timestamp": "2020-12-07T11:07:08.000Z", "user": { "id": "l7gk7f82" }, "message": "Logout successful" }
----
You can now use the search API to run a URI search on this index.
The following URI search matches documents with a `user.id` value of `l7gk7f82`.
Note the query is specified using the `q` query string parameter.
[source,console]
----
GET /user_logs_000001/_search?q=user.id:8a4f500d
----
// TEST[continued]
The API returns the following response. Note the `hits.hits` property contains
the document that matched the query.
[source,console-result]
----
{
"took": 2,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 1,
"relation": "eq"
},
"max_score": 0.9808291,
"hits": [
{
"_index": "user_logs_000001",
"_type": "_doc",
"_id": "2",
"_score": 0.9808291,
"_source": {
"@timestamp": "2020-12-07T11:06:07.000Z",
"user": {
"id": "8a4f500d"
},
"message": "Login successful"
}
}
]
}
}
----
// TESTRESPONSE[s/"took": 2/"took": "$body.took"/]
====
[discrete]
[[run-request-body-search]]
=== Run a request body search
You can use the search API's <<request-body-search-query,`query` request
body parameter>> to provide a query as a JSON object, written in
<<query-dsl,Query DSL>>.
.*Example*
[%collapsible]
====
The following request body search uses the <<query-dsl-match-query,`match`>>
query to match documents with a `message` value of `login successful`. Note the
`match` query is specified as a JSON object in the `query` parameter.
[source,console]
----
GET /user_logs_000001/_search
{
"query": {
"match": {
"message": "login successful"
}
}
}
----
// TEST[continued]
The API returns the following response.
The `hits.hits` property contains matching documents. By default, the response
sorts these matching documents by `_score`, a <<relevance-scores,relevance
score>> that measures how well each document matches the query.
[source,console-result]
----
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 3,
"relation": "eq"
},
"max_score": 0.9983525,
"hits": [
{
"_index": "user_logs_000001",
"_type": "_doc",
"_id": "2",
"_score": 0.9983525,
"_source": {
"@timestamp": "2020-12-07T11:06:07.000Z",
"user": {
"id": "8a4f500d"
},
"message": "Login successful"
}
},
{
"_index": "user_logs_000001",
"_type": "_doc",
"_id": "3",
"_score": 0.49917626,
"_source": {
"@timestamp": "2020-12-07T11:07:08.000Z",
"user": {
"id": "l7gk7f82"
},
"message": "Logout successful"
}
},
{
"_index": "user_logs_000001",
"_type": "_doc",
"_id": "1",
"_score": 0.42081726,
"_source": {
"@timestamp": "2020-12-06T11:04:05.000Z",
"user": {
"id": "vlb44hny"
},
"message": "Login attempt failed"
}
}
]
}
}
----
// TESTRESPONSE[s/"took": 1/"took": "$body.took"/]
====
[discrete]
[[search-multiple-indices]]
=== Search multiple indices
To search multiple indices, add them as comma-separated values in the search API
request path.
.*Example*
[%collapsible]
====
The following request searches the `user_logs_000001` and `user_logs_000002`
indices.
[source,console]
----
GET /user_logs_000001,user_logs_000002/_search
{
"query": {
"match": {
"message": "login successful"
}
}
}
----
// TEST[continued]
// TEST[s/^/PUT user_logs_000002\n/]
====
You can also search multiple indices using an index pattern.
.*Example*
[%collapsible]
====
The following request uses the index pattern `user_logs*` in place of the index
name. The request searches any indices in the cluster that start with
`user_logs`.
[source,console]
----
GET /user_logs*/_search
{
"query": {
"match": {
"message": "login successful"
}
}
}
----
// TEST[continued]
====
To search all indices in a cluster, omit the index name from the request path.
Alternatively, you can use `_all` or `*` in place of the index name.
.*Example*
[%collapsible]
====
The following requests are equivalent and search all indices in the cluster.
[source,console]
----
GET /_search
{
"query": {
"match": {
"message": "login successful"
}
}
}
GET /_all/_search
{
"query": {
"match": {
"message": "login successful"
}
}
}
GET /*/_search
{
"query" : {
"match" : { "message" : "login" }
}
}
----
// TEST[continued]
====
include::request/from-size.asciidoc[]