Skywalking 存储模型

发表于 分类于 阅读次数：

这篇文章讲述 Skywalking 的存储模型。

概述

在 Skywalking 中主要存储有四种类型的数据：

  1. metric: 时序数据，该数据的特点是跟时间强相关，一般有四个元素构成：1) 名称；2）时间戳；3）数值；4）标签(tag),表示对指标的属性；
  2. trace: 调用链数据；
  3. log: 日志类数据，如项目中 logback 等框架输出的日志；
  4. event: 事件类型，如服务启动。

其界面如下所示：

1. Metric
skywalking-dashboard

2. Trace
skywalking-trace

3. Log
skywalking-log

4. Event
skywalking-event

Service-Instance-Endpoint 概念

在 Skywalking 中，有三个比较重要的概念：1）Service; 2) Instance; 3) Endpoint, 在 Metric, Trace, Log 数据中都会关联这些属性，从而方便统计和查询：

  • Service: 代表一个可以提供服务能力的实体，可对应微服务体系中的一个服务；
  • Instance: 代表一个服务实例进程，一个 Service 可以运行多个 Instance；
  • Endpoint: 代表一个请求接口，一个服务中可以包含多个 Endpoint.

工程结构

假定有如下的服务，其中 NacosProviderApp 是服务提供方，提供了一个 /hello/{string} endpoint, NacosConsumerApp 是服务消费方，通过浏览器访问。

skywalking-example

Service-Instance-Endpoint 索引

项目接入 Skywalking 中，会使用相应的索引存储这些信息。

索引名称 描述
{namespace}_service_traffic-yyyyMMdd 服务信息
{namespace}_instance_traffic-yyyyMMdd 服务实例信息
{namespace}_endpoint_traffic-yyyyMMdd Endpoint 信息
{namespace}_service_relation_server_side-yyyyMMdd 服务之间的关联关系

说明：

  1. 假定 namespace 为 my-elasticsearch;
  2. 每一份表会根据业务情况按天生成索引。

Service 索引

1. Mapping
通过命令 GET my-elasticsearch_service_traffic-20220701/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
{
  "my-elasticsearch_service_traffic-20220701" : {
    "mappings" : {
      "properties" : {
        "name" : {
          "type" : "keyword",
          "copy_to" : [
            "name_match"
          ]
        },
        "name_match" : {
          "type" : "text",
          "analyzer" : "oap_analyzer"
        },
        "node_type" : {
          "type" : "integer"
        },
        "service_group" : {
          "type" : "keyword"
        },
        "time_bucket" : {
          "type" : "long"
        }
      }
    }
  }
}

2. 索引结构

字段名称 类型 备注
_id keyword 唯一id
name keyword 服务名称
node_type integer
service_group keyword 服务分组
time_bucket long 时间桶

3. 数据
通过命令 GET my-elasticsearch_service_traffic-20220701/_search 查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 2,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "my-elasticsearch_service_traffic-20220701",
        "_type" : "_doc",
        "_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
        "_score" : 1.0,
        "_source" : {
          "node_type" : 0,
          "name" : "NacosConsumerApp",
          "service_group" : null
        }
      },
      {
        "_index" : "my-elasticsearch_service_traffic-20220701",
        "_type" : "_doc",
        "_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
        "_score" : 1.0,
        "_source" : {
          "node_type" : 0,
          "name" : "NacosProviderApp",
          "service_group" : null
        }
      }
    ]
  }
}
_id name node_type service_group time_bucket
TmFjb3NQcm92aWRlckFwcA==.1 NacosProviderApp 0 null
TmFjb3NDb25zdW1lckFwcA==.1 NacosConsumerApp 0 null

Instance 索引

1. Mapping
通过命令 GET my-elasticsearch_instance_traffic-20220701/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
{
  "my-elasticsearch_instance_traffic-20220701" : {
    "mappings" : {
      "properties" : {
        "last_ping" : {
          "type" : "long"
        },
        "name" : {
          "type" : "keyword",
          "index" : false
        },
        "properties" : {
          "type" : "text",
          "index" : false
        },
        "service_id" : {
          "type" : "keyword"
        },
        "time_bucket" : {
          "type" : "long"
        }
      }
    }
  }
}

2. 索引结构

字段名称 类型 备注
_id keyword 唯一id
last_ping long 最近一次 ping 的时间
name keyword “index” : false
properties text 属性字段，不索引，”index” : false
service_id keyword
time_bucket long 时间桶

3. 数据
通过命令 GET my-elasticsearch_instance_traffic-20220701/_search 查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 4,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "my-elasticsearch_instance_traffic-20220701",
        "_type" : "_doc",
        "_id" : "TmFjb3NDb25zdW1lckFwcA==.1_ZWI4MjlmMzQ5ZjE4NGQ3NmI2OGQwZWI2MGNjYmFiYzFAMTkyLjE2OC42OC42Mg==",
        "_score" : 1.0,
        "_source" : {
          "last_ping" : 202207021756,
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "name" : "eb829f349f184d76b68d0eb60ccbabc1@192.168.68.62",
          "time_bucket" : 202207011826,
          "properties" : """{"OS Name":"Windows 10","hostname":"DESKTOP-ICO161","Process No.":"7640","language":"java","ipv4s":"192.168.68.62"}"""
        }
      },
      {
        "_index" : "my-elasticsearch_instance_traffic-20220701",
        "_type" : "_doc",
        "_id" : "TmFjb3NQcm92aWRlckFwcA==.1_ZGMzMDZkYTI1NWVkNDJjMWIxYjU2NmNlZTI3NDYwZWRAMTkyLjE2OC42OC42Mg==",
        "_score" : 1.0,
        "_source" : {
          "last_ping" : 202207021757,
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "name" : "dc306da255ed42c1b1b566cee27460ed@192.168.68.62",
          "time_bucket" : 202207011827,
          "properties" : """{"OS Name":"Windows 10","hostname":"DESKTOP-ICO161","Process No.":"22100","language":"java","ipv4s":"192.168.68.62"}"""
        }
      }
    ]
  }
}

_id last_ping name service_id properties time_bucket
Tm… 202207021756 {..}@192.168.68.62 TmFjb3NDb25zdW1lckFwcA==.1 … 202207011826
Tm… 202207021757 {...}@192.168.68.62 TmFjb3NQcm92aWRlckFwcA==.1 … 202207011827

说明：

  • 通过 last_ping 字段进行服务的保活；
  • 通过 service_id 关联服务索引；
  • time_bucket 字段记录了服务启动的时间；
  • service_id 标识了一个服务实例，格式为：{随机字串}@hostname(ip).

Endpoint 索引

1. Mapping
通过命令 GET my-elasticsearch_endpoint_traffic-20220626/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
{
  "my-elasticsearch_endpoint_traffic-20220626" : {
    "mappings" : {
      "properties" : {
        "name" : {
          "type" : "keyword",
          "copy_to" : [
            "name_match"
          ]
        },
        "name_match" : {
          "type" : "text",
          "analyzer" : "oap_analyzer"
        },
        "service_id" : {
          "type" : "keyword"
        },
        "time_bucket" : {
          "type" : "long"
        }
      }
    }
  }
}

2. 索引结构

字段名称 类型 备注
_id keyword 唯一id
name keyword 将内容 copy_to name_match字段，进行分词索引
name_match text 进行分词索引
service_id keyword 关联到服务索引
time_bucket long 时间桶

3. 数据

通过命令 GET my-elasticsearch_endpoint_traffic-20220626/_search 查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 2,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "my-elasticsearch_endpoint_traffic-20220626",
        "_type" : "_doc",
        "_id" : "TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
        "_score" : 1.0,
        "_source" : {
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "name" : "{GET}/echo/{str}",
          "time_bucket" : 202206261908
        }
      },
      {
        "_index" : "my-elasticsearch_endpoint_traffic-20220626",
        "_type" : "_doc",
        "_id" : "TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
        "_score" : 1.0,
        "_source" : {
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "name" : "{GET}/services/hello/{str}",
          "time_bucket" : 202206261908
        }
      }
    ]
  }
}

_id name service_id time_bucket
Tm… {GET}/echo/{str} TmFjb3NDb25zdW1lckFwcA==.1 202206261908
Tm… {GET}/services/hello/{str} TmFjb3NQcm92aWRlckFwcA==.1 202206261908

服务关系索引

1. Mapping
通过命令 GET my-elasticsearch_service_relation_server_side-20220627/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
{
  "my-elasticsearch_service_relation_server_side-20220627" : {
    "mappings" : {
      "properties" : {
        "component_id" : {
          "type" : "integer",
          "index" : false
        },
        "dest_service_id" : {
          "type" : "keyword"
        },
        "entity_id" : {
          "type" : "keyword"
        },
        "source_service_id" : {
          "type" : "keyword"
        },
        "time_bucket" : {
          "type" : "long"
        }
      }
    }
  }
}

2. 索引结构

字段名称 类型 备注
_id keyword 唯一id
component_id integer
source_service_id keyword 源服务 id
dest_service_id keyword 目标服务 id
entity_id keyword
time_bucket long 时间桶

3. 数据
通过命令 GET my-elasticsearch_service_relation_server_side-20220627/_search 查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 2,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "my-elasticsearch_service_relation_server_side-20220627",
        "_type" : "_doc",
        "_id" : "202206271817_TmFjb3NDb25zdW1lckFwcA==.1-TmFjb3NQcm92aWRlckFwcA==.1",
        "_score" : 1.0,
        "_source" : {
          "source_service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "component_id" : 14,
          "dest_service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "time_bucket" : 202206271817,
          "entity_id" : "TmFjb3NDb25zdW1lckFwcA==.1-TmFjb3NQcm92aWRlckFwcA==.1"
        }
      },
      {
        "_index" : "my-elasticsearch_service_relation_server_side-20220627",
        "_type" : "_doc",
        "_id" : "202206271817_VXNlcg==.0-TmFjb3NDb25zdW1lckFwcA==.1",
        "_score" : 1.0,
        "_source" : {
          "source_service_id" : "VXNlcg==.0",
          "component_id" : 14,
          "dest_service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "time_bucket" : 202206271817,
          "entity_id" : "VXNlcg==.0-TmFjb3NDb25zdW1lckFwcA==.1"
        }
      }
    ]
  }
}



| _id | component_id | source_service_id | dest_service_id | entity_id | time_bucket |
| ———– | ———– | ———– | ———– |
| 202.. | 14 | TmFjb3NDb25zdW1lckFwcA==.1 | TmFjb3NQcm92aWRlckFwcA==.1 | Tm… | 202206271817 |
| 202.. | 14 | VXNlcg==.0 | TmFjb3NDb25zdW1lckFwcA==.1 | VX… | 202206271817 |

说明：
通过 source_service_id 和 dest_service_id 两个字段，可以建立起服务之间的调用关系，从而得到服务之间的拓扑图。根据上面的关系，可以得出拓扑图为：VXNlcg==.0(user) --> TmFjb3NDb25zdW1lckFwcA==.1(NacosConsumerApp) --> TmFjb3NQcm92aWRlckFwcA==.1(NacosProviderApp).

skywalking-topology

Metric

Skywalking 可以存储以下类别的 Metric 数据：APM, Database, Event, Istio, Istio Data Plane, K8s, SelfObservability, VM, Web Browser. 项目中对接 Skywalking 之后，便可采集 APM 的数据。在本文中，我们分析 APM metric 指标，下图展示了 APM 中包含的所有指标。

skywalking-apm-metric

统计维度

APM metric 指标可以从四个不同的维度进行统计，分别是 Global, Service, Instance, Endpoint. 其含义如下：Global 是从全局整体的维度统计数据，其它维度参考上文的内容。

指标 索引 描述
all_percentile {namespace}_metrics-percentile-yyyyMMdd 服务健康程度
service_apdex {namespace}_metrics-apdex-yyyyMMdd 服务健康程度
service_resp_time {namespace}_metrics-longavg-yyyyMMdd 服务平均响应时间
service_sla {namespace}_metrics-percent-yyyyMMdd 服务请求成功率
service_cpm {namespace}_metrics-cpm-yyyyMMdd 服务每分钟请求数
service_percentile {namespace}_metrics-percentile-yyyyMMdd 百分比响应延时
service_instance_cpm {namespace}_metrics-cpm-yyyyMMdd 服务实例每分钟请求数
service_instance_resp_time {namespace}_metrics-longavg-yyyyMMdd 每个服务实例的最大延时
service_instance_sla {namespace}_metrics-percent-yyyyMMdd 每个服务实例的请求成功率
instance_jvm_cpu {namespace}_metrics-doubleavg-yyyyMMdd jvm占用CPU的百分比
instance_jvm_memory_heap {namespace}_metrics-longavg-yyyyMMdd JVM内存占用大小，单位m
instance_jvm_young_gc_time {namespace}_metrics-sum-yyyyMMdd JVM垃圾回收时间
instance_jvm_thread_live_count {namespace}_metrics-longavg-yyyyMMdd JVM线程数量
instance_jvm_thread_runnable_state_thread_count {namespace}_metrics-longavg-yyyyMMdd JVM不同状态线程数量
instance_jvm_class_loaded_class_count {namespace}_metrics-longavg-yyyyMMdd JVM对象的数量
endpoint_cpm {namespace}_metrics-cpm-yyyyMMdd 每个端点的每分钟请求数
endpoint_avg {namespace}_metrics-longavg-yyyyMMdd 每个端点的最慢请求时间，单位ms
endpoint_sla {namespace}_metrics-percent-yyyyMMdd 每个端点的请求成功率
endpoint_percentile {namespace}_metrics-percentile-yyyyMMdd 当前端点每个时间段的响应时间占比

下面以 endpoint_avg, service_resp_time,endpoint_percentile 为例，分析其存储模型。

endpoint_avg: 请求响应的平均时间

1. Mapping
通过命令 GET my-elasticsearch_metrics-longavg-20220702/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
{
  "my-elasticsearch_metrics-longavg-20220702" : {
    "mappings" : {
      "properties" : {
        "count" : {
          "type" : "long",
          "index" : false
        },
        "entity_id" : {
          "type" : "keyword"
        },
        "metric_table" : {
          "type" : "keyword"
        },
        "service_id" : {
          "type" : "keyword"
        },
        "summation" : {
          "type" : "long",
          "index" : false
        },
        "time_bucket" : {
          "type" : "long"
        },
        "value" : {
          "type" : "long"
        }
      }
    }
  }
}


2. 索引结构

字段名称 类型 备注
_id keyword 唯一id
count long 数量
entity_id keyword 根据 metric_table 值不同，代表不同的含义，在这里，是指 endpont Id 值
metric_table keyword 指标名称
service_id keyword service Id
summation long 总和
value long 平均值
time_bucket long 时间桶, 这里是S, 如 202207021443

说明：

  • metrics-longavg 索引了值为 long 类型的指标；
  • metric_table 字段存储了指标的名称；
  • entity_id 字段，根据 metric_table 值不同，代表不同的含义，在这里，是指 endpont Id 值；
  • service_id 字段存储了 service 的 id;
  • endpoint_avg 指标记录了 endpoint 每秒请求的数量、平均值及总和。

3. 数据
通过以下命令查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
POST /my-elasticsearch_metrics-longavg-20220702/_search
{
  "query": {
    "bool": {
      "must": [{
        "term": {
          "metric_table": {
            "value": "endpoint_avg",
            "boost": 1.0
          }
        }
      }]
    }
  }
}

结果如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 4,
      "relation" : "eq"
    },
    "max_score" : 9.878767,
    "hits" : [
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_avg_202207021443_TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
        "_score" : 9.878767,
        "_source" : {
          "metric_table" : "endpoint_avg",
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "count" : 1,
          "time_bucket" : 202207021443,
          "entity_id" : "TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
          "value" : 468,
          "summation" : 468
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_avg_202207021443_TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
        "_score" : 9.878767,
        "_source" : {
          "metric_table" : "endpoint_avg",
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "count" : 1,
          "time_bucket" : 202207021443,
          "entity_id" : "TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
          "value" : 514,
          "summation" : 514
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_avg_202207021444_TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
        "_score" : 9.878767,
        "_source" : {
          "metric_table" : "endpoint_avg",
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "count" : 8,
          "time_bucket" : 202207021444,
          "entity_id" : "TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
          "value" : 518,
          "summation" : 4145
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_avg_202207021444_TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
        "_score" : 9.878767,
        "_source" : {
          "metric_table" : "endpoint_avg",
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "count" : 8,
          "time_bucket" : 202207021444,
          "entity_id" : "TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
          "value" : 521,
          "summation" : 4171
        }
      }
    ]
  }
}

_id metric_table service_id entity_id count summation value time_bucket
endpoint… endpoint_avg TmFjb3NQcm92aWRlckFwcA==.1 TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0= 1 468 468 202207021443
endpoint… endpoint_avg TmFjb3NDb25zdW1lckFwcA==.1 TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ== 1 4145 514 202207021443
endpoint… endpoint_avg TmFjb3NQcm92aWRlckFwcA==.1 TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0= 8 4145 518 202207021444
endpoint… endpoint_avg TmFjb3NDb25zdW1lckFwcA==.1 TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ== 8 4171 521 202207021444

第一行表示在 202207021443 S, NacosProviderApp 服务的 {GET}/services/hello/{str} endpint 请求了一次，平均值为 468 ms;
第一行表示在 202207021443 S, NacosConsumerApp 服务的 {GET}/echo/{str} endpint 请求了一次，平均值为 514 ms;

service_resp_time: 服务平均响应时间

1. Mapping

Mapping 见上文；

2. 索引结构

索引结构见上文。

说明：

  • entity_id 字段在这里代表了 service id.

3. 数据
通过以下命令查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
POST /my-elasticsearch_metrics-longavg-20220702/_search
{
  "query": {
    "bool": {
      "must": [{
        "term": {
          "metric_table": {
            "value": "service_resp_time",
            "boost": 1.0
          }
        }
      }]
    }
  }
}

结果如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 4,
      "relation" : "eq"
    },
    "max_score" : 9.861144,
    "hits" : [
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "service_resp_time_202207021443_TmFjb3NDb25zdW1lckFwcA==.1",
        "_score" : 9.861144,
        "_source" : {
          "metric_table" : "service_resp_time",
          "count" : 1,
          "time_bucket" : 202207021443,
          "entity_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "value" : 514,
          "summation" : 514
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "service_resp_time_202207021443_TmFjb3NQcm92aWRlckFwcA==.1",
        "_score" : 9.861144,
        "_source" : {
          "metric_table" : "service_resp_time",
          "count" : 1,
          "time_bucket" : 202207021443,
          "entity_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "value" : 468,
          "summation" : 468
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "service_resp_time_202207021444_TmFjb3NDb25zdW1lckFwcA==.1",
        "_score" : 9.861144,
        "_source" : {
          "metric_table" : "service_resp_time",
          "count" : 8,
          "time_bucket" : 202207021444,
          "entity_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "value" : 521,
          "summation" : 4171
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-longavg-20220702",
        "_type" : "_doc",
        "_id" : "service_resp_time_202207021444_TmFjb3NQcm92aWRlckFwcA==.1",
        "_score" : 9.861144,
        "_source" : {
          "metric_table" : "service_resp_time",
          "count" : 8,
          "time_bucket" : 202207021444,
          "entity_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "value" : 518,
          "summation" : 4145
        }
      }
    ]
  }
}


_id metric_table service_id entity_id count summation value time_bucket
service… service_resp_time TmFjb3NQcm92aWRlckFwcA==.1 1 468 468 202207021443
service… service_resp_time TmFjb3NDb25zdW1lckFwcA==.1 1 514 514 202207021443
service… service_resp_time TmFjb3NQcm92aWRlckFwcA==.1 8 4145 518 202207021444
service… service_resp_time TmFjb3NDb25zdW1lckFwcA==.1 8 4171 521 202207021444

第一行表示在 202207021443 S, NacosProviderApp 服务被请求了一次，平均值为 468 ms;
第一行表示在 202207021443 S, NacosConsumerApp 服务被请求了一次，平均值为 514 ms;

可见 service_resp_time 指标是指服务所有 endpoint 每秒请求的平均值、总数和总和。由于服务只有一个 endpoint, 所以，它与 endpoint_avg 指标值相同。

endpoint_percentile: endpoint 每个时间段的响应时间占比

1. Mapping
通过命令 GET my-elasticsearch_metrics-percentile-20220702/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
{
  "my-elasticsearch_metrics-percentile-20220702" : {
    "mappings" : {
      "properties" : {
        "dataset" : {
          "type" : "text",
          "index" : false
        },
        "entity_id" : {
          "type" : "keyword"
        },
        "metric_table" : {
          "type" : "keyword"
        },
        "precision" : {
          "type" : "integer",
          "index" : false
        },
        "service_id" : {
          "type" : "keyword"
        },
        "time_bucket" : {
          "type" : "long"
        },
        "value" : {
          "type" : "text",
          "index" : false
        }
      }
    }
  }
}


2. 索引结构

字段名称 类型 备注
_id keyword 唯一id
dataset text 数据集
entity_id keyword 根据 metric_table 值不同，代表不同的含义
metric_table keyword 指标名称
service_id keyword service Id
precision integer
value text 指标值
time_bucket long 时间桶, 这里是S, 如 202207021443

3. 数据
通过以下命令查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
POST /my-elasticsearch_metrics-percentile-20220702/_search
{
  "query": {
    "bool": {
      "must": [{
        "term": {
          "metric_table": {
            "value": "endpoint_percentile",
            "boost": 1.0
          }
        }
      }]
    }
  }
}

结果如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 4,
      "relation" : "eq"
    },
    "max_score" : 1.8170772,
    "hits" : [
      {
        "_index" : "my-elasticsearch_metrics-percentile-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_percentile_202207021443_TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
        "_score" : 1.8170772,
        "_source" : {
          "metric_table" : "endpoint_percentile",
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "precision" : 10,
          "time_bucket" : 202207021443,
          "entity_id" : "TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
          "value" : "0,460|1,460|2,460|3,460|4,460",
          "dataset" : "46,1"
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-percentile-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_percentile_202207021443_TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
        "_score" : 1.8170772,
        "_source" : {
          "metric_table" : "endpoint_percentile",
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "precision" : 10,
          "time_bucket" : 202207021443,
          "entity_id" : "TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
          "value" : "0,510|1,510|2,510|3,510|4,510",
          "dataset" : "51,1"
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-percentile-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_percentile_202207021444_TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
        "_score" : 1.8170772,
        "_source" : {
          "metric_table" : "endpoint_percentile",
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "precision" : 10,
          "time_bucket" : 202207021444,
          "entity_id" : "TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
          "value" : "0,530|1,750|2,780|3,940|4,940",
          "dataset" : "67,1|78,1|35,1|2,1|7,1|94,1|75,1|53,1"
        }
      },
      {
        "_index" : "my-elasticsearch_metrics-percentile-20220702",
        "_type" : "_doc",
        "_id" : "endpoint_percentile_202207021444_TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
        "_score" : 1.8170772,
        "_source" : {
          "metric_table" : "endpoint_percentile",
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "precision" : 10,
          "time_bucket" : 202207021444,
          "entity_id" : "TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
          "value" : "0,540|1,760|2,780|3,940|4,940",
          "dataset" : "67,1|78,1|35,1|3,1|7,1|94,1|76,1|54,1"
        }
      }
    ]
  }
}


_id metric_table service_id entity_id precision dataset value time_bucket
endpoint… endpoint_percentile TmFjb3NQcm92aWRlckFwcA==.1 TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0= 10 … … 202207021443
endpoint… endpoint_percentile TmFjb3NDb25zdW1lckFwcA==.1 TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ== 10 … … 202207021443
endpoint… endpoint_percentile TmFjb3NQcm92aWRlckFwcA==.1 TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0= 10 … … 202207021444
endpoint… endpoint_percentile TmFjb3NDb25zdW1lckFwcA==.1 TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ== 10 … … 202207021444

第一行表示在 202207021443 S, NacosProviderApp 服务的 {GET}/services/hello/{str} endpint 的 endpoint_percentile 值为 0,460|1,460|2,460|3,460|4,460;
第一行表示在 202207021443 S, NacosConsumerApp 服务的 {GET}/echo/{str} endpint 的 endpoint_percentile 值为 0,510|1,510|2,510|3,510|4,510.

说明：

  • 字段 dataset 和 value 存储的值不理解，后续再深入理解。

Trace

Trace 日志存储在 {namespace}_segment-yyyyMMdd 索引中。

Mapping

通过命令 GET my-elasticsearch_segment-20220702/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
{
  "my-elasticsearch_segment-20220702" : {
    "mappings" : {
      "properties" : {
        "data_binary" : {
          "type" : "binary"
        },
        "end_time" : {
          "type" : "long"
        },
        "endpoint_id" : {
          "type" : "keyword"
        },
        "endpoint_name" : {
          "type" : "keyword",
          "copy_to" : [
            "endpoint_name_match"
          ]
        },
        "endpoint_name_match" : {
          "type" : "text",
          "analyzer" : "oap_analyzer"
        },
        "is_error" : {
          "type" : "integer"
        },
        "latency" : {
          "type" : "integer"
        },
        "segment_id" : {
          "type" : "keyword"
        },
        "service_id" : {
          "type" : "keyword"
        },
        "service_instance_id" : {
          "type" : "keyword"
        },
        "start_time" : {
          "type" : "long"
        },
        "statement" : {
          "type" : "keyword"
        },
        "tags" : {
          "type" : "keyword"
        },
        "time_bucket" : {
          "type" : "long"
        },
        "trace_id" : {
          "type" : "keyword"
        },
        "version" : {
          "type" : "integer",
          "index" : false
        }
      }
    }
  }
}


索引结构

字段名称 类型 备注
_id keyword 唯一id
trace_id keyword trace id
segment_id keyword segment id
service_id keyword service id
service_instance_id keyword service instance id
endpoint_id keyword endpoint id
endpoint_name keyword endpoint name
endpoint_name_match text 对 endpoint name 分词索引
tags keyword 标签，可根据标签进行 Keyword 匹配查询
statement keyword statement
is_error integer 是否错误
latency integer 延时
version integer version
tags keyword 标签，可根据标签进行 Keyword 匹配查询
start_time long 开始时间
end_time long 结束时间
data_binary data_binary 完整数据，通过Base64编码存储
time_bucket long 时间桶

说明：

  • tags 字段：使用数组存储，可以动态添加多个 tag, 但只能使用内置 tag, 不能自定义 tag;
  • data_binary 字段：segment 内容通过 Base64 编码存储，不能索引。

数据

通过以下命令 GET my-elasticsearch_segment-20220702/_search 查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
{
  "took" : 1,
  "timed_out" : false,
  "_shards" : {
    "total" : 5,
    "successful" : 5,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 18,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "my-elasticsearch_segment-20220702",
        "_type" : "_doc",
        "_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.88.16567442415110000",
        "_score" : 1.0,
        "_source" : {
          "trace_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.88.16567442415110001",
          "endpoint_name" : "{GET}/echo/{str}",
          "latency" : 673,
          "end_time" : 1656744242184,
          "endpoint_id" : "TmFjb3NDb25zdW1lckFwcA==.1_e0dFVH0vZWNoby97c3RyfQ==",
          "service_instance_id" : "TmFjb3NDb25zdW1lckFwcA==.1_ZWI4MjlmMzQ5ZjE4NGQ3NmI2OGQwZWI2MGNjYmFiYzFAMTkyLjE2OC42OC42Mg==",
          "version" : 3,
          "tags" : [
            "http.method=GET",
            "http.method=GET"
          ],
          "start_time" : 1656744241511,
          "data_binary" : "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",
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "statement" : "{GET}/echo/{str} - 5227fe6d8fbc4af3871f4b5d6c0a78ba.88.16567442415110001",
          "time_bucket" : 20220702144401,
          "is_error" : 0,
          "segment_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.88.16567442415110000"
        }
      },
      {
        "_index" : "my-elasticsearch_segment-20220702",
        "_type" : "_doc",
        "_id" : "4dcbfc1f16324d7fabf5bf2d2d12bb8a.85.16567442453060000",
        "_score" : 1.0,
        "_source" : {
          "trace_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.89.16567442453030001",
          "endpoint_name" : "{GET}/services/hello/{str}",
          "latency" : 756,
          "end_time" : 1656744246062,
          "endpoint_id" : "TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
          "service_instance_id" : "TmFjb3NQcm92aWRlckFwcA==.1_ZGMzMDZkYTI1NWVkNDJjMWIxYjU2NmNlZTI3NDYwZWRAMTkyLjE2OC42OC42Mg==",
          "version" : 3,
          "tags" : [
            "http.method=GET"
          ],
          "start_time" : 1656744245306,
          "data_binary" : "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",
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "statement" : "{GET}/services/hello/{str} - 5227fe6d8fbc4af3871f4b5d6c0a78ba.89.16567442453030001",
          "time_bucket" : 20220702144405,
          "is_error" : 0,
          "segment_id" : "4dcbfc1f16324d7fabf5bf2d2d12bb8a.85.16567442453060000"
        }
      }
      
      // ...
    ]
  }
}

为了简化，只复制了部分数据。一条完成的 Segment 数据如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
{
  "trace_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.89.16567442453030001",
  "endpoint_name" : "{GET}/services/hello/{str}",
  "latency" : 756,
  "end_time" : 1656744246062,
  "endpoint_id" : "TmFjb3NQcm92aWRlckFwcA==.1_e0dFVH0vc2VydmljZXMvaGVsbG8ve3N0cn0=",
  "service_instance_id" : "TmFjb3NQcm92aWRlckFwcA==.1_ZGMzMDZkYTI1NWVkNDJjMWIxYjU2NmNlZTI3NDYwZWRAMTkyLjE2OC42OC42Mg==",
  "version" : 3,
  "tags" : [
    "http.method=GET"
  ],
  "start_time" : 1656744245306,
  "data_binary" : "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",
  "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
  "statement" : "{GET}/services/hello/{str} - 5227fe6d8fbc4af3871f4b5d6c0a78ba.89.16567442453030001",
  "time_bucket" : 20220702144405,
  "is_error" : 0,
  "segment_id" : "4dcbfc1f16324d7fabf5bf2d2d12bb8a.85.16567442453060000"
}

Log

Log 日志存储在 {namespace}_log-yyyyMMdd 索引中。

Mapping

通过命令 GET my-elasticsearch_log-20220702/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
{
  "my-elasticsearch_log-20220702" : {
    "mappings" : {
      "properties" : {
        "content" : {
          "type" : "keyword",
          "copy_to" : [
            "content_match"
          ]
        },
        "content_match" : {
          "type" : "text",
          "analyzer" : "oap_log_analyzer"
        },
        "content_type" : {
          "type" : "integer",
          "index" : false
        },
        "endpoint_id" : {
          "type" : "keyword"
        },
        "endpoint_name" : {
          "type" : "keyword",
          "copy_to" : [
            "endpoint_name_match"
          ]
        },
        "endpoint_name_match" : {
          "type" : "text",
          "analyzer" : "oap_analyzer"
        },
        "service_id" : {
          "type" : "keyword"
        },
        "service_instance_id" : {
          "type" : "keyword"
        },
        "span_id" : {
          "type" : "integer"
        },
        "tags" : {
          "type" : "keyword"
        },
        "tags_raw_data" : {
          "type" : "binary"
        },
        "time_bucket" : {
          "type" : "long"
        },
        "timestamp" : {
          "type" : "long"
        },
        "trace_id" : {
          "type" : "keyword"
        },
        "trace_segment_id" : {
          "type" : "keyword"
        },
        "unique_id" : {
          "type" : "keyword"
        }
      }
    }
  }
}


索引结构

字段名称 类型 备注
_id keyword 唯一id
trace_id keyword trace id
trace_segment_id keyword trace segment id
unique_id keyword unique id
span_id integer span id
service_id keyword service id
service_instance_id keyword service instance id
endpoint_id keyword endpoint id
endpoint_name keyword endpoint name
endpoint_name_match text 对 endpoint name 分词索引
tags keyword 标签，可根据标签进行 Keyword 匹配查询
tags_raw_data binary 标签原始内容，使用 Base64 编码存储
statement keyword statement
is_error integer 是否错误
latency integer 延时
version integer version
tags keyword 标签，可根据标签进行 Keyword 匹配查询
timestamp long 时间戳
content keyword 日志内容
content_match text 日志内容
content_type integer content type
content keyword 对日志内容分词索引
time_bucket long 时间桶

说明：

  • tags 字段：使用数组存储，可以动态添加多个 tag, 但只能使用内置 tag, 不能自定义 tag, 该字段只存储被索引的字段;
  • tags_raw_data 字段：tags 原始内容通过 Base64 编码存储，不能索引；

数据

通过以下命令 GET my-elasticsearch_segment-20220702/_search 查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
{
  "took" : 1,
  "timed_out" : false,
  "_shards" : {
    "total" : 5,
    "successful" : 5,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 18,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "my-elasticsearch_log-20220702",
        "_type" : "_doc",
        "_id" : "fd0b1e10acaf4e089cec6b402a348cb2",
        "_score" : 1.0,
        "_source" : {
          "trace_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.87.16567442400710001",
          "unique_id" : "fd0b1e10acaf4e089cec6b402a348cb2",
          "span_id" : 0,
          "endpoint_name" : null,
          "endpoint_id" : null,
          "service_instance_id" : "TmFjb3NQcm92aWRlckFwcA==.1_ZGMzMDZkYTI1NWVkNDJjMWIxYjU2NmNlZTI3NDYwZWRAMTkyLjE2OC42OC42Mg==",
          "content" : """2022-07-02 14:44:00.075 [TID:5227fe6d8fbc4af3871f4b5d6c0a78ba.87.16567442400710001] [http-nio-8070-exec-6] INFO  o.n.n.NacosProviderApplication$EchoController -Receive a request:skywalking
""",
          "tags" : [
            "level=INFO"
          ],
          "trace_segment_id" : "4dcbfc1f16324d7fabf5bf2d2d12bb8a.83.16567442400740000",
          "content_type" : 1,
          "tags_raw_data" : "Cg0KBWxldmVsEgRJTkZPCkQKBmxvZ2dlchI6b3JnLm5vYWhzYXJrLm5hY29zLk5hY29zUHJvdmlkZXJBcHBsaWNhdGlvbiRFY2hvQ29udHJvbGxlcgoeCgZ0aHJlYWQSFGh0dHAtbmlvLTgwNzAtZXhlYy02",
          "service_id" : "TmFjb3NQcm92aWRlckFwcA==.1",
          "time_bucket" : 20220702144400,
          "timestamp" : 1656744240075
        }
      },
      {
        "_index" : "my-elasticsearch_log-20220702",
        "_type" : "_doc",
        "_id" : "826963aef56b4bbab4f607c46b43faeb",
        "_score" : 1.0,
        "_source" : {
          "trace_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.80.16567442485970003",
          "unique_id" : "826963aef56b4bbab4f607c46b43faeb",
          "span_id" : 0,
          "endpoint_name" : null,
          "endpoint_id" : null,
          "service_instance_id" : "TmFjb3NDb25zdW1lckFwcA==.1_ZWI4MjlmMzQ5ZjE4NGQ3NmI2OGQwZWI2MGNjYmFiYzFAMTkyLjE2OC42OC42Mg==",
          "content" : """2022-07-02 14:44:08.598 [TID:5227fe6d8fbc4af3871f4b5d6c0a78ba.80.16567442485970003] [http-nio-9092-exec-1] INFO  o.n.n.NacosConsumerApplication$TestController -Receive a request:skywalking
""",
          "tags" : [
            "level=INFO"
          ],
          "trace_segment_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.80.16567442485970002",
          "content_type" : 1,
          "tags_raw_data" : "Cg0KBWxldmVsEgRJTkZPCkQKBmxvZ2dlchI6b3JnLm5vYWhzYXJrLm5hY29zLk5hY29zQ29uc3VtZXJBcHBsaWNhdGlvbiRUZXN0Q29udHJvbGxlcgoeCgZ0aHJlYWQSFGh0dHAtbmlvLTkwOTItZXhlYy0x",
          "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
          "time_bucket" : 20220702144408,
          "timestamp" : 1656744248598
        }
      }
      // ...
    ]
  }
}


为了简化，只复制了部分数据。一条完成的 log 数据如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
{
  "trace_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.80.16567442485970003",
  "unique_id" : "826963aef56b4bbab4f607c46b43faeb",
  "span_id" : 0,
  "endpoint_name" : null,
  "endpoint_id" : null,
  "service_instance_id" : "TmFjb3NDb25zdW1lckFwcA==.1_ZWI4MjlmMzQ5ZjE4NGQ3NmI2OGQwZWI2MGNjYmFiYzFAMTkyLjE2OC42OC42Mg==",
  "content" : """2022-07-02 14:44:08.598 [TID:5227fe6d8fbc4af3871f4b5d6c0a78ba.80.16567442485970003] [http-nio-9092-exec-1] INFO  o.n.n.NacosConsumerApplication$TestController -Receive a request:skywalking""",
  "tags" : [
    "level=INFO"
  ],
  "trace_segment_id" : "5227fe6d8fbc4af3871f4b5d6c0a78ba.80.16567442485970002",
  "content_type" : 1,
  "tags_raw_data" : "Cg0KBWxldmVsEgRJTkZPCkQKBmxvZ2dlchI6b3JnLm5vYWhzYXJrLm5hY29zLk5hY29zQ29uc3VtZXJBcHBsaWNhdGlvbiRUZXN0Q29udHJvbGxlcgoeCgZ0aHJlYWQSFGh0dHAtbmlvLTkwOTItZXhlYy0x",
  "service_id" : "TmFjb3NDb25zdW1lckFwcA==.1",
  "time_bucket" : 20220702144408,
  "timestamp" : 1656744248598
}

Event

Event 日志存储在 {namespace}_events-yyyyMMdd 索引中。

Mapping

通过命令 GET my-elasticsearch_events-20220701/_mapping 查看索引结构。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
{
  "my-elasticsearch_events-20220701" : {
    "mappings" : {
      "properties" : {
        "end_time" : {
          "type" : "long"
        },
        "endpoint" : {
          "type" : "keyword"
        },
        "message" : {
          "type" : "keyword"
        },
        "name" : {
          "type" : "keyword"
        },
        "parameters" : {
          "type" : "keyword",
          "index" : false
        },
        "service" : {
          "type" : "keyword"
        },
        "service_instance" : {
          "type" : "keyword"
        },
        "start_time" : {
          "type" : "long"
        },
        "time_bucket" : {
          "type" : "long"
        },
        "type" : {
          "type" : "keyword"
        },
        "uuid" : {
          "type" : "keyword"
        }
      }
    }
  }
}

索引结构

字段名称 类型 备注
_id keyword 唯一id
uuid keyword uuid
service keyword service
service_instance keyword service instance
endpoint keyword endpoint
name keyword name
type keyword type
start_time long 开始时间
end_time long 结束时间
message keyword 事件内容
parameters keyword 启动参数
time_bucket long 时间桶

说明：

  • service 字段：服务名称;
  • service_instance 字段：服务实例名称；

数据

通过以下命令 GET my-elasticsearch_events-20220701/_search 查看数据。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
{
  "took" : 0,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 4,
      "relation" : "eq"
    },
    "max_score" : 1.0,
    "hits" : [
      {
        "_index" : "my-elasticsearch_events-20220701",
        "_type" : "_doc",
        "_id" : "09e90a94-9154-48a5-a32c-6c2e18ca695d",
        "_score" : 1.0,
        "_source" : {
          "start_time" : 1656671199452,
          "endpoint" : "",
          "service" : "NacosConsumerApp",
          "name" : "Start",
          "end_time" : 1656671200887,
          "time_bucket" : 202207011826,
          "service_instance" : "",
          "type" : "Normal",
          "message" : "Start Java Application",
          "uuid" : "09e90a94-9154-48a5-a32c-6c2e18ca695d",
          "parameters" : """{"OPTS":"-Dcom.sun.management.jmxremote -Dcom.sun.management.jmxremote.authenticate\u003dfalse -Dcom.sun.management.jmxremote.port\u003d62148 -Dcom.sun.management.jmxremote.ssl\u003dfalse -Dfile.encoding\u003dUTF-8 -Djava.rmi.server.hostname\u003dlocalhost -Dspring.application.admin.enabled\u003dtrue -Dspring.liveBeansView.mbeanDomain -Dspring.output.ansi.enabled\u003dalways -XX:TieredStopAtLevel\u003d1 -Xverify:none -javaagent:C:/Java/skywalking-es7-8.6.0/agent/skywalking-agent.jar -javaagent:C:\\Java\\IDEA-2018.3.2\\lib\\idea_rt.jar\u003d62149:C:\\Java\\IDEA-2018.3.2\\bin"}"""
        }
      },
      {
        "_index" : "my-elasticsearch_events-20220701",
        "_type" : "_doc",
        "_id" : "3c64b57a-8f1c-4cd0-9622-fb9a665936b6",
        "_score" : 1.0,
        "_source" : {
          "start_time" : 1656671244223,
          "endpoint" : "",
          "service" : "NacosConsumerApp",
          "name" : "Start",
          "end_time" : 1656671245361,
          "time_bucket" : 202207011827,
          "service_instance" : "",
          "type" : "Normal",
          "message" : "Start Java Application",
          "uuid" : "3c64b57a-8f1c-4cd0-9622-fb9a665936b6",
          "parameters" : """{"OPTS":"-Dcom.sun.management.jmxremote -Dcom.sun.management.jmxremote.authenticate\u003dfalse -Dcom.sun.management.jmxremote.port\u003d62207 -Dcom.sun.management.jmxremote.ssl\u003dfalse -Dfile.encoding\u003dUTF-8 -Djava.rmi.server.hostname\u003dlocalhost -Dspring.application.admin.enabled\u003dtrue -Dspring.liveBeansView.mbeanDomain -Dspring.output.ansi.enabled\u003dalways -XX:TieredStopAtLevel\u003d1 -Xverify:none -javaagent:C:/Java/skywalking-es7-8.6.0/agent/skywalking-agent.jar -javaagent:C:\\Java\\IDEA-2018.3.2\\lib\\idea_rt.jar\u003d62208:C:\\Java\\IDEA-2018.3.2\\bin"}"""
        }
      }

      // ...
    ]
  }
}

为了简化，只复制了部分数据。一条完成的 event 数据如下：

1
2
3
4
5
6
7
8
9
10
11
12
13
14
{
  "start_time" : 1656671244223,
  "endpoint" : "",
  "service" : "NacosConsumerApp",
  "name" : "Start",
  "end_time" : 1656671245361,
  "time_bucket" : 202207011827,
  "service_instance" : "",
  "type" : "Normal",
  "message" : "Start Java Application",
  "uuid" : "3c64b57a-8f1c-4cd0-9622-fb9a665936b6",
  "parameters" : """{"OPTS":"-Dcom.sun.management.jmxremote -Dcom.sun.management.jmxremote.authenticate\u003dfalse -Dcom.sun.management.jmxremote.port\u003d62207 -Dcom.sun.management.jmxremote.ssl\u003dfalse -Dfile.encoding\u003dUTF-8 -Djava.rmi.server.hostname\u003dlocalhost -Dspring.application.admin.enabled\u003dtrue -Dspring.liveBeansView.mbeanDomain -Dspring.output.ansi.enabled\u003dalways -XX:TieredStopAtLevel\u003d1 -Xverify:none -javaagent:C:/Java/skywalking-es7-8.6.0/agent/skywalking-agent.jar -javaagent:C:\\Java\\IDEA-2018.3.2\\lib\\idea_rt.jar\u003d62208:C:\\Java\\IDEA-2018.3.2\\bin"}"""
}

参考：

1. SkyWalking调研与初步实践
2. skywalking UI指标说明