ELK+grok如何收集並格式化MySQL的slow日誌？

前言

需求說明

用ELK收集MySQL的慢查詢日誌信息，並且使用grok插件將日誌信息格式化爲json格式。

部署安排

test101——10.0.0.101——部署MySQL、filebeat
test102——10.0.0.102——部署elasticsearch、kibana
test103——10.0.0.103——部署logstash

實驗環境版本

系統版本：centos7.3
數據庫版本：

本次測試了3個版本數據庫
1、yum安裝的mariadb5.5（Server version: 5.5.60-MariaDB MariaDB Server）
2、yum安裝的mariadb10.4（Server version: 10.4.0-MariaDB-log MariaDB Server）
3、源碼安裝的MySQL5.7.20

ELK版本：6.5.3

操作步驟（MariaDB5.5.60爲例）

1、安裝mariadb，準備數據庫環境

1.1 安裝服務

[root@test101 ~]# yum -y install mariadb*     #這樣安裝默認是5.5.60版本  

1.2 修改配置文件，開啓慢查詢日誌

在/etc/my.conf文件加入下面三行內容，開啓慢查詢日誌：

slow_query_log
long_query_time = 2      #闕值設置成2秒，只要超過2秒的操作，就會寫入slow-log文件
slow_query_log_file = "/var/lib/mysql/test101-slow.log"    #設置慢查詢日誌

重啓MySQL

[root@test101 ~]# systemctl restart mysql 

1.3 創建測試庫，導入數據

1）先創建一個大數據量的文件：

[root@test101 ~]# seq 1 19999999 > /tmp/big

2）登錄數據庫創建測試數據庫：

[root@test101 ~]# mysql
MariaDB [(none)]> create database test101;     #創建測試庫
MariaDB [(none)]> use test101;
MariaDB [test101]> create table table1(id int(10)not null)engine=innodb;    #創建測試表

3）將剛剛創建好的大數據文件/tmp/big導入到MySQL的測試數據庫中，生成測試數據表：

MariaDB [test101]> show tables;
+----------------------+
| Tables_in_test101 |
+----------------------+
| table1                   |
+-----------------------+
1 row in set (0.00 sec)

MariaDB [test101]> load data local infile '/tmp/big' into table table1;   #注意：local不能少，少了local可能會報錯“ERROR 13 (HY000): Can't get stat of '/tmp/big' (Errcode: 2)”
Query OK, 19999999 rows affected (1 min 47.64 sec)   
Records: 19999999  Deleted: 0  Skipped: 0  Warnings: 0

MariaDB [test101]> 

測試數據導入成功

1.4 生成慢查詢日誌

在數據庫查詢一條數據：

MariaDB [test101]> select * from table1 where id=258;
+-----+
| id  |
+-----+
| 258 |
+-----+
1 row in set (11.76 sec)

MariaDB [test101]> 

然後查看慢查詢日誌文件，已經生成了相應的日誌。數據庫準備完成：

[root@test101 mysql]# tailf /var/lib/mysql/test101-slow.log 

# Time: 181217 15:23:39
# User@Host: root[root] @ localhost []
# Thread_id: 2  Schema: test101  QC_hit: No
# Query_time: 11.758867  Lock_time: 0.000106  Rows_sent: 1  Rows_examined: 19999999
SET timestamp=1545031419;
select * from table1 where id=258;

2、ELK工具部署

2.1 在test101安裝好filebeat、test102安裝elasticsearch/kibana

（安裝和基礎配置省略）

2.2 在test103服務器安裝好logstash

（安裝步驟和基礎配置省略）

1）創建logstash採集日誌文件：

[root@test103 conf.d]# cat /etc/logstash/conf.d/logstash-syslog.conf 
# Sample Logstash configuration for creating a simple
# Beats -> Logstash -> Elasticsearch pipeline.

input {
  beats {
    port => 5044
  }
}

output {
  elasticsearch {
    hosts => ["http://10.0.0.102:9200"]
    index => "%{[@metadata][beat]}-%{[@metadata][version]}-%{+YYYY.MM.dd}"
    #user => "elastic"
    #password => "changeme"
  }
  stdout {
    codec => rubydebug
  }

}

2.3 啓動服務查看日誌採集

啓動101服務器的filebeat，102服務器的elasticsearch、前臺啓動103服務器的logstash，前臺查看日誌採集

1）在test101服務器的MySQL運行查詢語句（注意，運行查詢語句之前要保證logstash前臺啓動已經完成）：

MariaDB [test101]> select * from table1 where id=358;

+-----+
| id  |
+-----+
| 358 |
+-----+
1 row in set (13.47 sec)

MariaDB [test101]> 

2）查看test101服務器上慢查詢日誌已經生成：

[root@test101 mysql]# tailf test101-slow.log 

# Time: 181218  8:58:30
# User@Host: root[root] @ localhost []
# Thread_id: 3  Schema: test101  QC_hit: No
# Query_time: 13.405630  Lock_time: 0.000271  Rows_sent: 1  Rows_examined: 19999999
SET timestamp=1545094710;
select * from table1 where id=358;

3）再查看logstash，已經採集到相關信息（但是上面的6句日誌信息，是被分成了6段messages採集的）：

[root@test103 logstash]# /usr/share/logstash/bin/logstash -f /etc/logstash/conf.d/logstash-syslog.conf
.......#省略若干行啓動信息
[INFO ] 2018-12-18 08:51:48.669 [Api Webserver] agent - Successfully started Logstash API endpoint {:port=>9600}  
#就是出現這個提示的時候才運行上面的mysql查詢語句

{
       "message" => "# Time: 181218  8:58:30",
    "@timestamp" => 2018-12-18T00:58:33.396Z,
    "prospector" => {
        "type" => "log"
    },
         "input" => {
        "type" => "log"
    },
      "@version" => "1",
          "host" => {
                   "os" => {
            "codename" => "Core",
            "platform" => "centos",
             "version" => "7 (Core)",
              "family" => "redhat"
        },
                 "name" => "test101",
        "containerized" => true,
         "architecture" => "x86_64",
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc"
    },
          "beat" => {
            "name" => "test101",
        "hostname" => "test101",
         "version" => "6.5.3"
    },
          "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
        "offset" => 252,
        "source" => "/var/lib/mysql/test101-slow.log"
}
{
       "message" => "# User@Host: root[root] @ localhost []",
    "@timestamp" => 2018-12-18T00:58:33.398Z,
    "prospector" => {
        "type" => "log"
    },
         "input" => {
        "type" => "log"
    },
      "@version" => "1",
          "host" => {
        "containerized" => true,
                 "name" => "test101",
         "architecture" => "x86_64",
                   "os" => {
            "codename" => "Core",
            "platform" => "centos",
             "version" => "7 (Core)",
              "family" => "redhat"
        },
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc"
    },
          "beat" => {
            "name" => "test101",
         "version" => "6.5.3",
        "hostname" => "test101"
    },
          "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
        "source" => "/var/lib/mysql/test101-slow.log",
        "offset" => 276
}
{
       "message" => "# Thread_id: 3  Schema: test101  QC_hit: No",
    "@timestamp" => 2018-12-18T00:58:33.398Z,
    "prospector" => {
        "type" => "log"
    },
         "input" => {
        "type" => "log"
    },
      "@version" => "1",
          "host" => {
                   "os" => {
              "family" => "redhat",
            "platform" => "centos",
             "version" => "7 (Core)",
            "codename" => "Core"
        },
                 "name" => "test101",
        "containerized" => true,
         "architecture" => "x86_64",
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc"
    },
          "beat" => {
            "name" => "test101",
        "hostname" => "test101",
         "version" => "6.5.3"
    },
          "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
        "source" => "/var/lib/mysql/test101-slow.log",
        "offset" => 315
}
{
       "message" => "# Query_time: 13.405630  Lock_time: 0.000271  Rows_sent: 1  Rows_examined: 19999999",
    "@timestamp" => 2018-12-18T00:58:33.398Z,
    "prospector" => {
        "type" => "log"
    },
         "input" => {
        "type" => "log"
    },
      "@version" => "1",
          "host" => {
                 "name" => "test101",
         "architecture" => "x86_64",
                   "os" => {
            "codename" => "Core",
            "platform" => "centos",
             "version" => "7 (Core)",
              "family" => "redhat"
        },
        "containerized" => true,
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc"
    },
          "beat" => {
            "name" => "test101",
        "hostname" => "test101",
         "version" => "6.5.3"
    },
          "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
        "source" => "/var/lib/mysql/test101-slow.log",
        "offset" => 359
}
{
       "message" => "SET timestamp=1545094710;",
    "@timestamp" => 2018-12-18T00:58:33.398Z,
    "prospector" => {
        "type" => "log"
    },
         "input" => {
        "type" => "log"
    },
      "@version" => "1",
          "host" => {
        "containerized" => true,
                 "name" => "test101",
         "architecture" => "x86_64",
                   "os" => {
            "codename" => "Core",
            "platform" => "centos",
             "version" => "7 (Core)",
              "family" => "redhat"
        },
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc"
    },
          "beat" => {
            "name" => "test101",
        "hostname" => "test101",
         "version" => "6.5.3"
    },
          "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
        "source" => "/var/lib/mysql/test101-slow.log",
        "offset" => 443
}
{
       "message" => "select * from table1 where id=358;",
    "@timestamp" => 2018-12-18T00:58:33.398Z,
    "prospector" => {
        "type" => "log"
    },
         "input" => {
        "type" => "log"
    },
      "@version" => "1",
          "host" => {
         "architecture" => "x86_64",
                   "os" => {
            "codename" => "Core",
            "platform" => "centos",
             "version" => "7 (Core)",
              "family" => "redhat"
        },
                 "name" => "test101",
        "containerized" => true,
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc"
    },
          "beat" => {
            "name" => "test101",
         "version" => "6.5.3",
        "hostname" => "test101"
    },
          "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
        "source" => "/var/lib/mysql/test101-slow.log",
        "offset" => 469
}

3、日誌採集優化

因爲慢日誌裏面，一條信息被分成了6段，不便於日誌的分析，因此需要將上面採集到的6個messages優化合併成一個

3.1 修改filebeat配置，將上面的6段messages合併成一句日誌信息

修改filebeat配置文件，在日誌路徑後面加入下面三行，然後重啓filebeat：

  paths:
    - /var/lib/mysql/test101-slow.log
    #- c:\programdata\elasticsearch\logs\*
    #加入下面三行
  multiline.pattern: "^# User@Host:"
  multiline.negate: true
  multiline.match: after

說明：

multiline.pattern：正則表達式，去匹配指定的一行，這裏去匹配的以“# User@Host:”開頭的那一行；
multiline.negate：true 或 false；默認是false，就是將multiline.pattern匹配到的那一行合併到上一行；如果配置是true，就是將除了multiline.pattern匹的那一行的其他所有行合併到其上一行；
multiline.match：after 或 before，就是指定將要合併到上一行的內容，合併到上一行的末尾或開頭；

3.2 檢查修改後的採集日誌格式

1）再在MySQL執行一次查詢語句：

MariaDB [test101]> select * from table1 where id=368;
+-----+
| id  |
+-----+
| 368 |
+-----+
1 row in set (12.54 sec)

MariaDB [test101]> 

2）查看看查詢日誌，已經生成日誌：

[root@test101 mysql]# tailf test101-slow.log 

# Time: 181218  9:17:42
# User@Host: root[root] @ localhost []
# Thread_id: 3  Schema: test101  QC_hit: No
# Query_time: 12.541603  Lock_time: 0.000493  Rows_sent: 1  Rows_examined: 19999999
SET timestamp=1545095862;
select * from table1 where id=368;

3）查看logstash的實時輸出，採集到的日誌就變成了一行：

[root@test103 logstash]# /usr/share/logstash/bin/logstash -f /etc/logstash/conf.d/logstash-syslog.conf
.......#省略若干行啓動信息
{
       "message" => # User@Host: root[root] @ localhost []\n# Thread_id: 3  Schema: test101  QC_hit: No\n# Query_time: 12.541603  Lock_time: 0.000493  Rows_sent: 1  Rows_examined: 19999999\nSET timestamp=1545095862;\nselect * from table1 where id=368;",   
                   #這一行就是在slow-log中的信息，已經被合併成了一行，慢日誌的“"# Time: 181218  9:17:42” 單獨成行，在前面沒有貼出來
    "prospector" => {
        "type" => "log"
    },
         "input" => {
        "type" => "log"
    },
          "host" => {
         "architecture" => "x86_64",
                   "os" => {
            "codename" => "Core",
            "platform" => "centos",
             "version" => "7 (Core)",
              "family" => "redhat"
        },
        "containerized" => true,
                 "name" => "test101",
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc"
    },
           "log" => {
        "flags" => [
            [0] "multiline"
        ]
    },
    "@timestamp" => 2018-12-18T01:17:45.336Z,
      "@version" => "1",
          "beat" => {
            "name" => "test101",
        "hostname" => "test101",
         "version" => "6.5.3"
    },
          "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
        "source" => "/var/lib/mysql/test101-slow.log",
        "offset" => 504
}

4、logstash配置grok插件，格式化分析日誌

在實現了上一步驟，即將日誌的messages合併之後，就需要採用grok插件將messages格式化成json格式了。

4.1 編輯logstash-syslog.conf文件配置grok插件

編輯logstash-syslog.conf文件，在input和output中間，加入整個filter模塊，配置如下：

[root@test103 conf.d]# cat  logstash-syslog.conf
# Sample Logstash configuration for creating a simple
# Beats -> Logstash -> Elasticsearch pipeline.

input {
  beats {
    port => 5044
  }
}

filter {

   #這一步格式化messages爲json格式
    grok {
       match => [ "message", "(?m)^# User@Host: %{USER:query_user}\[[^\]]+\] @ (?:(?<query_host>\S*) )?\[(?:%{IP:query_ip})?\]\s# Thread_id:\s+%{NUMBER:thread_id:int}\s+Schema: %{USER:schema}\s+QC_hit: %{WORD:QC_hit}\s*# Query_time: %{NUMBER:query_time:float}\s+Lock_time: %{NUMBER:lock_time:float}\s+Rows_sent: %{NUMBER:rows_sent:int}\s+Rows_examined: %{NUMBER:rows_examined:int}\s*(?:use %{DATA:database};\s*)?SET timestamp=%{NUMBER:timestamp};\s*(?<query>(?<action>\w+)\s+.*)" ]

       }
#這一步是將日誌中的時間那一行（如：# Time: 181218  9:17:42）加上一個“drop”的tag
    grok {
        match => { "message" => "# Time: " }
        add_tag => [ "drop" ]
        tag_on_failure => []
    }
#刪除標籤中含有drop的行。也就是要刪除慢日誌裏面的“# Time: 181218  9:17:42”這樣的內容
    if "drop" in [tags] {
        drop {}
    }

#   時間轉換
    date {
        match => ["mysql.slowlog.timestamp", "UNIX", "YYYY-MM-dd HH:mm:ss"]
        target => "@timestamp"
        timezone => "Asia/Shanghai"
    }

    ruby {
        code => "event.set('[@metadata][today]', Time.at(event.get('@timestamp').to_i).localtime.strftime('%Y.%m.%d'))"
    }

#刪除字段message
    mutate {
        remove_field => [ "message" ]
    }
}

output {
  elasticsearch {
    hosts => ["http://10.0.0.102:9200"]
    #index =>  "%{tags[0]}" 
    index => "%{[@metadata][beat]}-%{[@metadata][version]}-%{+YYYY.MM.dd}"
    #user => "elastic"
    #password => "changeme"
  }
  stdout {
    codec => rubydebug
  }

}

4.2 檢查格式化後的日誌採集

1）清除elasticsearch的舊索引，保證索引乾淨：

[root@test102 filebeat]# curl 10.0.0.102:9200/_cat/indices
green open .kibana_1 udOUvbprSnKWUJISwD0r_g 1 0 8 0 74.4kb 74.4kb
[root@test102 filebeat]# 

2）重啓filebeat，重新啓動logstash前臺輸出，在mariadb執行查詢語句：

MariaDB [test101]> select * from table1 where id=588;
+-----+
| id  |
+-----+
| 588 |
+-----+
1 row in set (13.00 sec)

MariaDB [test101]> 

3）查看慢查詢日誌已經生成：

[root@test101 mysql]# tailf test101-slow.log 

# Time: 181218 14:39:38
# User@Host: root[root] @ localhost []
# Thread_id: 4  Schema: test101  QC_hit: No
# Query_time: 12.999487  Lock_time: 0.000303  Rows_sent: 1  Rows_examined: 19999999
SET timestamp=1545115178;
select * from table1 where id=588;

4）查看logstash的實時採集輸出，已經是json格式了：

[INFO ] 2018-12-18 14:25:02.802 [Api Webserver] agent - Successfully started Logstash API endpoint {:port=>9600}

{
           "QC_hit" => "No",
        "rows_sent" => 1,
           "action" => "select",
            "input" => {
        "type" => "log"
    },
           "offset" => 1032,
         "@version" => "1",
       "query_host" => "localhost",
             "beat" => {
            "name" => "test101",
        "hostname" => "test101",
         "version" => "6.5.3"
    },
            "query" => "select * from table1 where id=428;",
       "@timestamp" => 2018-12-18T06:25:34.017Z,
       "prospector" => {
        "type" => "log"
    },
        "lock_time" => 0.000775,
       "query_user" => "root",
              "log" => {
        "flags" => [
            [0] "multiline"
        ]
    },
           "schema" => "test101",
        "timestamp" => "1545114330",
        "thread_id" => 4,
       "query_time" => 14.155133,
             "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
             "host" => {
                 "name" => "test101",
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc",
        "containerized" => true,
                   "os" => {
            "codename" => "Core",
              "family" => "redhat",
             "version" => "7 (Core)",
            "platform" => "centos"
        },
         "architecture" => "x86_64"
    },
    "rows_examined" => 19999999,
           "source" => "/var/lib/mysql/test101-slow.log"
{
           "QC_hit" => "No",
        "rows_sent" => 1,
           "action" => "select",
            "input" => {
        "type" => "log"
    },
           "offset" => 1284,
         "@version" => "1",
       "query_host" => "localhost",
             "beat" => {
            "name" => "test101",
        "hostname" => "test101",
         "version" => "6.5.3"
    },
            "query" => "select * from table1 where id=588;",
       "@timestamp" => 2018-12-18T06:39:44.082Z,
       "prospector" => {
        "type" => "log"
    },
        "lock_time" => 0.000303,
       "query_user" => "root",
              "log" => {
        "flags" => [
            [0] "multiline"
        ]
    },
           "schema" => "test101",
        "timestamp" => "1545115178",
        "thread_id" => 4,
       "query_time" => 12.999487,
             "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
             "host" => {
                 "name" => "test101",
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc",
        "containerized" => true,
                   "os" => {
            "codename" => "Core",
              "family" => "redhat",
             "version" => "7 (Core)",
            "platform" => "centos"
        },
         "architecture" => "x86_64"
    },
    "rows_examined" => 19999999,
           "source" => "/var/lib/mysql/test101-slow.log"
}

再打開kibana界面，建立索引，然後查看kibana的界面，日誌OK：

其他版本MySQL慢日誌及grok語句

1、MySQL5.7.20版本

mysql5.7.20版本慢日誌：

# Time: 2018-12-18T08:43:24.828892Z
# User@Host: root[root] @ localhost []  Id:     7
# Query_time: 15.819314  Lock_time: 0.000174 Rows_sent: 1  Rows_examined: 19999999
SET timestamp=1545122604;
select * from table1 where id=258;

grok語法：

match => [ "message", "(?m)^# User@Host: %{USER:query_user}\[[^\]]+\] @ (?:(?<query_host>\S*) )?\[(?:%{IP:query_ip})?\]\s+Id:\s+%{NUMBER:id:int}\s# Query_time: %{NUMBER:query_time:float}\s+Lock_time: %{NUMBER:lock_time:float}\s+Rows_sent: %{NUMBER:rows_sent:int}\s+Rows_examined: %{NUMBER:rows_examined:int}\s*(?:use %{DATA:database};\s*)?SET timestamp=%{NUMBER:timestamp};\s*(?<query>(?<action>\w+)\s+.*)" ]

logstash採集查詢結果：

{
    "rows_examined" => 19999999,
            "input" => {
        "type" => "log"
    },
             "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
       "query_user" => "root",
       "query_time" => 15.461892,
       "query_host" => "localhost",
         "@version" => "1",
            "query" => "select * from table1 where id=258;",
             "host" => {
        "containerized" => true,
                   "os" => {
            "platform" => "centos",
             "version" => "7 (Core)",
              "family" => "redhat",
            "codename" => "Core"
        },
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc",
                 "name" => "test101",
         "architecture" => "x86_64"
    },
           "source" => "/home/data/test101-mysql5.7.20-slow.log",
           "action" => "select",
        "lock_time" => 0.000181,
       "prospector" => {
        "type" => "log"
    },
           "offset" => 1416,
              "log" => {
        "flags" => [
            [0] "multiline"
        ]
    },
        "rows_sent" => 1,
        "timestamp" => "1545122934",
             "beat" => {
        "hostname" => "test101",
         "version" => "6.5.3",
            "name" => "test101"
    },
       "@timestamp" => 2018-12-18T08:48:58.531Z,
               "id" => 7
}

2、MariaDB10.4版本

mariadb10.4配置慢查詢和5.5.60不同，是在/etc/my.cnf.d/server.cnf文件的[mysqld]模塊下添加配置開啓慢查詢，不是改/etc/my.conf:

[root@test101 filebeat]# cat /etc/my.cnf.d/server.cnf 
......   #省略若干行
[mysqld]
 slow_query_log
 long_query_time = 2
 slow_query_log_file = "/var/lib/mysql/test101-slow.log"
......   #省略若干行

mariadb10.4版本慢日誌：

# Time: 181219 13:01:17
# User@Host: root[root] @ localhost []
# Thread_id: 8  Schema: test101  QC_hit: No
# Query_time: 18.711204  Lock_time: 0.000443  Rows_sent: 1  Rows_examined: 19999999
# Rows_affected: 0  Bytes_sent: 80     #比5.5.60版本多了 Rows_affected和 Bytes_sent:字段
SET timestamp=1545195677;
select * from test101.table1 where id=888;

grok語法：

match => [ "message", "(?m)^# User@Host: %{USER:query_user}\[[^\]]+\] @ (?:(?<query_host>\S*) )?\[(?:%{IP:query_ip})?\]\s# Thread_id:\s+%{NUMBER:thread_id:int}\s+Schema: %{USER:schema}\s+QC_hit: %{WORD:QC_hit}\s*# Query_time: %{NUMBER:query_time:float}\s+Lock_time: %{NUMBER:lock_time:float}\s+Rows_sent: %{NUMBER:rows_sent:int}\s+Rows_examined: %{NUMBER:rows_examined:int}\s*# Rows_affected: %{NUMBER:rows_affected:int}\s+Bytes_sent: %{NUMBER:bytes_sent:int}\s*(?:use %{DATA:database};\s*)?SET timestamp=%{NUMBER:timestamp};\s*(?<query>(?<action>\w+)\s+.*)" ]

logstash採集查詢結果：

{
           "source" => "/var/lib/mysql/test101-slow.log",
         "@version" => "1",
       "query_time" => 18.711204,
           "offset" => 319,
       "query_user" => "root",
       "prospector" => {
        "type" => "log"
    },
           "schema" => "test101",
       "query_host" => "localhost",
             "tags" => [
        [0] "beats_input_codec_plain_applied"
    ],
           "action" => "select",
       "bytes_sent" => 80,
             "host" => {
                   "id" => "9fcb0c7f28e449aabad922d1ba1f0afc",
         "architecture" => "x86_64",
                   "os" => {
              "family" => "redhat",
            "codename" => "Core",
            "platform" => "centos",
             "version" => "7 (Core)"
        },
                 "name" => "test101",
        "containerized" => true
    },
    "rows_examined" => 19999999,
       "@timestamp" => 2018-12-19T05:01:23.310Z,
        "rows_sent" => 1,
            "query" => "select * from test101.table1 where id=888;",
              "log" => {
        "flags" => [
            [0] "multiline"
        ]
    },
        "thread_id" => 8,
             "beat" => {
        "hostname" => "test101",
            "name" => "test101",
         "version" => "6.5.3"
    },
           "QC_hit" => "No",
            "input" => {
        "type" => "log"
    },
    "rows_affected" => 0,
        "lock_time" => 0.000443,
        "timestamp" => "1545195677"
}