# \[elasticserach performance\] match query very slowly!
**URL:** https://discuss.elastic.co/t/elasticserach-performance-match-query-very-slowly/100743
**Category:** Elasticsearch
**Created:** [September 16, 2017, 2:38am UTC](https://discuss.elastic.co/t/elasticserach-performance-match-query-very-slowly/100743 "2017-09-16T02:38:55Z")
**Posts on this page:** 4
**Page:** 1
### Author:  [@laoyang360](https://discuss.elastic.co/u/laoyang360)
#### Post date: [September 16, 2017, 2:38am UTC](https://discuss.elastic.co/t/elasticserach-performance-match-query-very-slowly/100743/1 "2017-09-16T02:38:56Z")
1. Search the keyword "ladder" use matchQuery (java api) it will be very very very slowly!
Time consuming as follows:
> {
> "took": 7778,
> "timed\_out": false,
> "\_shards": {
> "total": 5,
> "successful": 5,
> "failed": 0
> },
> "hits": {
> "total": 92,
> "max\_score": 15.208916,
1. to do the matchquery full text search, the document is very large, the largest document about 2.5MB (return to the results)
2. Heap usage record, three nodes: 10% (master node), 4% (routing node), 58% (data node).
Want to analyze what causes it? And how to optimize the query?
----------------Dividing line-------------------------------- A
The following is profile monitoring records.
> "profile": {
> "shards": [
> {
> "id": "[BosI3A-bSJaVM0LHULDC9A] [baike\_index] [0]",
> "searches": [
> {
> "query": [
> {
> "type": "TermQuery",
> "description": "content: ladder"
> "time": "0.5679210000ms",
> "time\_in\_nanos": 567921,
> "breakdown": {
> "score": 4818,
> "build\_scorer\_count": 23,
> "match\_count": 0,
> "create\_weight": 490622,
> "next\_doc": 5347,
> "match": 0,
> "create\_weight\_count": 1,
> "next\_doc\_count": 21,
> "score\_count": 18,
> "build\_scorer": 67071,
> "advance": 0,
> "advance\_count": 0
> }
> }
> ],
> "rewrite\_time": 379627,
> "collector": [
> {
> "name": "CancellableCollector",
> "reason": "search\_cancelled",
> "time": "0.03252600000ms",
> "time\_in\_nanos": 32526,
> "children": [
> {
> "name": "SimpleTopScoreDocCollector",
> "reason": "search\_top\_hits",
> "time": "0.01902000000ms",
> "time\_in\_nanos": 19020
> }
> ]
> }
> ]
> }
> ],
> "aggregations":
> },
> {
> "id": "[BosI3A-bSJaVM0LHULDC9A] [baike\_index] [1]",
> "searches": [
> {
> "query": [
> {
> "type": "TermQuery",
> "description": "content: ladder"
> "time": "0.8072510000ms",
> "time\_in\_nanos": 807251,
> "breakdown": {
> "score": 5896,
> "build\_scorer\_count": 29,
> "match\_count": 0,
> "create\_weight": 745418,
> "next\_doc": 7775,
> "match": 0,
> "create\_weight\_count": 1,
> "next\_doc\_count": 16,
> "score\_count": 13,
> "build\_scorer": 48103,
> "advance": 0,
> "advance\_count": 0
> }
> }
> ],
> "rewrite\_time": 508630,
> "collector": [
> {
> "name": "CancellableCollector",
> "reason": "search\_cancelled",
> "time": "0.04867000000ms",
> "time\_in\_nanos": 48670,
> "children": [
> {
> "name": "SimpleTopScoreDocCollector",
> "reason": "search\_top\_hits",
> "time": "0.02597900000ms",
> "time\_in\_nanos": 25979
> }
> ]
> }
> ]
> }
> ],
> "aggregations":
> },
> {
> "id": "[BosI3A-bSJaVM0LHULDC9A] [baike\_index] [2]",
> "searches": [
> {
> "query": [
> {
> "type": "TermQuery",
> "description": "content: ladder"
> "time": "0.8369420000ms",
> "time\_in\_nanos": 836942,
> "breakdown": {
> "score": 8774,
> "build\_scorer\_count": 31,
> "match\_count": 0,
> "create\_weight": 744262,
> "next\_doc": 11751,
> "match": 0,
> "create\_weight\_count": 1,
> "next\_doc\_count": 29,
> "score\_count": 25,
> "build\_scorer": 72069,
> "advance": 0,
> "advance\_count": 0
> }
> }
> ],
> "rewrite\_time": 458680,
> "collector": [
> {
> "name": "CancellableCollector",
> "reason": "search\_cancelled",
> "time": "0.06351200000ms",
> "time\_in\_nanos": 63512,
> "children": [
> {
> "name": "SimpleTopScoreDocCollector",
> "reason": "search\_top\_hits",
> "time": "0.03681100000ms",
> "time\_in\_nanos": 36811
> }
> ]
> }
> ]
> }
> ],
> "aggregations":
> },
> {
> "id": "[BosI3A-bSJaVM0LHULDC9A] [baike\_index] [4]",
> "searches": [
> {
> "query": [
> {
> "type": "TermQuery",
> "description": "content: ladder"
> "time": "0.9213890000ms",
> "time\_in\_nanos": 921389,
> "breakdown": {
> "score": 8899,
> "build\_scorer\_count": 28,
> "match\_count": 0,
> "create\_weight": 745401,
> "next\_doc": 11755,
> "match": 0,
> "create\_weight\_count": 1,
> "next\_doc\_count": 29,
> "score\_count": 22,
> "build\_scorer": 155254,
> "advance": 0,
> "advance\_count": 0
> }
> }
> ],
> "rewrite\_time": 508615,
> "collector": [
> {
> "name": "CancellableCollector",
> "reason": "search\_cancelled",
> "time": "0.07178900000ms",
> "time\_in\_nanos": 71789,
> "children": [
> {
> "name": "SimpleTopScoreDocCollector",
> "reason": "search\_top\_hits",
> "time": "0.04696100000ms",
> "time\_in\_nanos": 46961
> }
> ]
> }
> ]
> }
> ],
> "aggregations":
> },
> {
> "id": "[DbUASJSzTDWaLQocTYuKpA] [baike\_index] [3]",
> "searches": [
> {
> "query": [
> {
> "type": "TermQuery",
> "description": "content: ladder"
> "time": "0.3358920000ms",
> "time\_in\_nanos": 335892,
> "breakdown": {
> "score": 6083,
> "build\_scorer\_count": 26,
> "match\_count": 0,
> "create\_weight": 273080,
> "next\_doc": 4956,
> "match": 0,
> "create\_weight\_count": 1,
> "next\_doc\_count": 19,
> "score\_count": 14,
> "build\_scorer": 51713,
> "advance": 0,
> "advance\_count": 0
> }
> }
> ],
> "rewrite\_time": 265629,
> "collector": [
> {
> "name": "CancellableCollector",
> "reason": "search\_cancelled",
> "time": "0.02894000000ms",
> "time\_in\_nanos": 28940,
> "children": [
> {
> "name": "SimpleTopScoreDocCollector",
> "reason": "search\_top\_hits",
> "time": "0.01789500000ms",
> "time\_in\_nanos": 17895
> }
> ]
> }
> ]
> }
> ],
> "aggregations":
> }
> ]
> }
> }
---
### Author:  [@jpountz](https://discuss.elastic.co/u/jpountz)
#### Post date: [September 25, 2017, 8:40am UTC](https://discuss.elastic.co/t/elasticserach-performance-match-query-very-slowly/100743/2 "2017-09-25T08:40:22Z")
This is surprising indeed. The query is simple (TermQuery) and matches few documents. It should be very fast. Does it reproduce all the time?
---
### Author:  [@laoyang360](https://discuss.elastic.co/u/laoyang360)
#### Post date: [September 27, 2017, 8:10am UTC](https://discuss.elastic.co/t/elasticserach-performance-match-query-very-slowly/100743/3 "2017-09-27T08:10:44Z")
I already resolve it，because the document which i search is very very large and I use the plain method to highlight.
When I change the fvh highlight, the bug is go away.
Thanks!
---
### Author:  [@system](https://discuss.elastic.co/u/system)
#### Post date: [October 25, 2017, 8:11am UTC](https://discuss.elastic.co/t/elasticserach-performance-match-query-very-slowly/100743/4 "2017-10-25T08:11:04Z")
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.