# ORing a text field with a unique identifier keyword field leading to increased next\_doc count and poor performance **URL:** \ **Category:** Elasticsearch\ **Created:** [March 22, 2023, 6:11pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283 "2023-03-22T18:11:13Z")\ **Posts on this page:** 10\ **Page:** 1
**Author:** ![helderdias](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/helderdias/32/118779_2.png) [@helderdias](https://discuss.elastic.co/u/helderdias)\ **Post date:** [March 22, 2023, 6:11pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/1 "2023-03-22T18:11:13Z")
TL;DR: I want to find documents that match a certain query (e.g. "my search") OR match the unique ID of a document. When I search for the text field alone, the search is super fast. However, when I or the text field with the keyword field search becomes extremely slow. Looking at the profile data, it seems like the shard that contains the document ends up doing a lot more `next_doc` than the other shards. Why would that be the case? Below are the mappings and queries I executed for reference. Mapping: ```auto { "mappings": { "properties": { "searchable_name_1": { "type": "text" }, "searchable_name_2": { "type": "text" }, "unique_identifier": { "type": "keyword" } } } } ``` The following query executes super fast: ```auto { "profile": true, "query": { "bool": { "should": [ { "multi_match": { "query": "my search", "fields": [ "searchable_name_1", "searchable_name_2" ] } } ] } } } ``` ![Screenshot 2023-03-22 at 11.03.51 AM](https://us1.discourse-cdn.com/elastic/original/3X/6/b/6bed05b8d86c2a9440052b90d3944a108e7ba49d.png) However, when I or the unique identifier with the text query, it becomes super slow: ```auto { "profile": true, "query": { "bool": { "should": [ { "multi_match": { "query": "my search", "fields": [ "searchable_name_1", "searchable_name_2" ] } }, { "term": { "unique_identifier": { "value": "13590585", "boost": 10 } } } ] } } } ``` ![Screenshot 2023-03-22 at 11.02.44 AM](https://us1.discourse-cdn.com/elastic/original/3X/8/8/883f6c35964f1bd4fdd223c2cfb7bee74f9c7372.png) ![Screenshot 2023-03-22 at 11.12.02 AM](https://us1.discourse-cdn.com/elastic/original/3X/b/f/bff16a5bb27c5983a01bb73e3d4e5a5f587ed276.png) ---
**Author:** ![Mark\_Harwood1](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/mark_harwood1/32/101255_2.png) [@Mark\_Harwood1](https://discuss.elastic.co/u/Mark_Harwood1)\ **Post date:** [March 22, 2023, 7:36pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/2 "2023-03-22T19:36:24Z")
Hi Sanity check - the mapped field is called “unique\_identifier” and the query is “unique\_id” in your example ---
**Author:** ![helderdias](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/helderdias/32/118779_2.png) [@helderdias](https://discuss.elastic.co/u/helderdias)\ **Post date:** [March 22, 2023, 7:49pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/3 "2023-03-22T19:49:19Z")
Thanks for your reply. It was a copy-paste mistake while anonymizing the use case. Fixed in the original post. ---
**Author:** ![Mark\_Harwood1](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/mark_harwood1/32/101255_2.png) [@Mark\_Harwood1](https://discuss.elastic.co/u/Mark_Harwood1)\ **Post date:** [March 22, 2023, 8:30pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/4 "2023-03-22T20:30:49Z")
Thanks. How many docs have this unique ID of 13590585? ---
**Author:** ![helderdias](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/helderdias/32/118779_2.png) [@helderdias](https://discuss.elastic.co/u/helderdias)\ **Post date:** [March 22, 2023, 8:44pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/5 "2023-03-22T20:44:07Z")
Just one. That's why I'm confused why ORing the two queries would result in such latency since the posting list for that ID has only one doc. ---
**Author:** ![Mark\_Harwood1](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/mark_harwood1/32/101255_2.png) [@Mark\_Harwood1](https://discuss.elastic.co/u/Mark_Harwood1)\ **Post date:** [March 22, 2023, 9:03pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/6 "2023-03-22T21:03:23Z")
Hmm. Then I expect that the addition of a clause is causing a change to scoring that means [an optimisation](https://www.elastic.co/blog/faster-retrieval-of-top-hits-in-elasticsearch-with-block-max-wand) is not kicking in. If you set “track\_total\_hits” to true in both queries then this might help compare speeds with the shortcut optimisation disabled. (It won’t solve your problem but would help with the diagnostics here) ---
**Author:** ![helderdias](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/helderdias/32/118779_2.png) [@helderdias](https://discuss.elastic.co/u/helderdias)\ **Post date:** [March 22, 2023, 9:28pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/7 "2023-03-22T21:28:03Z")
I enabled `track_total_hits` on both queries. Fast one (just match query): 5323035 Slow one (match query OR keyword term): 5323036 ---
**Author:** ![Mark\_Harwood1](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/mark_harwood1/32/101255_2.png) [@Mark\_Harwood1](https://discuss.elastic.co/u/Mark_Harwood1)\ **Post date:** [March 22, 2023, 11:00pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/8 "2023-03-22T23:00:16Z")
I presume that’s the number of hits. How do the response times compare? ---
**Author:** ![helderdias](https://sea2.discourse-cdn.com/elastic/user_avatar/discuss.elastic.co/helderdias/32/118779_2.png) [@helderdias](https://discuss.elastic.co/u/helderdias)\ **Post date:** [March 22, 2023, 11:24pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/9 "2023-03-22T23:24:18Z")
That's correct. Shard response time for match only: ![Screenshot 2023-03-22 at 4.22.59 PM](https://us1.discourse-cdn.com/elastic/original/3X/8/c/8ca6db7fd6018534ac114026a69becef006080ac.png) Shard response time for match OR keyword term: ![Screenshot 2023-03-22 at 4.22.42 PM](https://us1.discourse-cdn.com/elastic/original/3X/9/5/95706d941fdb50362c530977fe9ca0cdb2b4bb63.png) By the way, thanks in advance for your help, Mark! ---
**Author:** ![system](https://us1.discourse-cdn.com/elastic/original/3X/1/a/1ac57faf039f6b580b3f104ef42a2a89e41014de.png) [@system](https://discuss.elastic.co/u/system)\ **Post date:** [April 19, 2023, 11:25pm UTC](https://discuss.elastic.co/t/oring-a-text-field-with-a-unique-identifier-keyword-field-leading-to-increased-next-doc-count-and-poor-performance/328283/10 "2023-04-19T23:25:10Z")
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.