TiDB新功能解读 DB和RG级别监控指标详细指南

网友投稿 255 2024-03-01



背景

在 TiDB 的 7.x 引入 record-db-label 新特性 https://github.com/pingcap/tidb/pull/41477,指标项添加 db 和 resource group 的 label,实现对 db 和资源组的指标监控,取代了 record-db-qps(存在已知的 bug)。

TiDB新功能解读 DB和RG级别监控指标详细指南

操作过程

配置和启用

tiup cluster edit-config {cluster-name},添加 record-db-label 配置项:

... server_configs: tidb: status.record-db-label: true ...

tiup cluster reload {cluster-name} -R tidb

mysql> show config where name like "%record-db-label%"; +------+------------------+------------------------+-------+ | Type | Instance | Name | Value | +------+------------------+------------------------+-------+ | tidb | 10.2.103.64:4101 | status.record-db-label | true |

查看数据

在 prometheus 的 web UI 上,可以查看 tidb_executor_statement_total、tidb_server_handle_query_duration_seconds、tidb_server_execute_error_total 是否已经有数据,并且有 db 和 resource group 的分类。

配置 grafana 指标项

新特性 record-db-label 在 tidb cloud serverless 上使用,从任一 serverless 集群的监控项上拷贝其公式。

简单修改后得到以下三条 grafana 公式

QPS Per DB sum(rate(tidb_executor_statement_total{k8s_cluster="$k8s_cluster", tidb_cluster="$tidb_cluster", instance=~"$instance", db!=""}[1m])) by (db,type) Average Query Duration Per DB sum(rate(tidb_server_handle_query_duration_seconds_sum{k8s_cluster="$k8s_cluster", tidb_cluster="$tidb_cluster", instance=~"$instance", db!=""}[1m])) by (db) / sum(rate(tidb_server_handle_query_duration_seconds_count{k8s_cluster="$k8s_cluster", tidb_cluster="$tidb_cluster", instance=~"$instance", db!=""}[1m])) by (db) Failed Query Per DB sum(rate(tidb_server_execute_error_total{k8s_cluster="$k8s_cluster", tidb_cluster="$tidb_cluster", instance=~"$instance", db!=""}[1m])) by (db)

编辑 Grafana 面板,vi /tidb-deploy/grafana-3000/bin/tidb.json

插入位置第 57 行

... "id": null, "iteration": 1655990780337, "links": [], "panels": [

插入内容

{ "collapsed": true, "datasource": null, "gridPos": { "h": 1, "w": 24, "x": 0, "y": 0 }, "id": 316, "panels": [ { "aliasColors": {}, "bars": false, "dashLength": 10, "dashes": false, "datasource": "${DS_TEST-CLUSTER}", "decimals": null, "description": "The number of SQL statements executed per second on every Database, which are collected by SQL types.", "editable": true, "error": false, "fieldConfig": { "defaults": {}, "overrides": [] }, "fill": 1, "fillGradient": 0, "grid": {}, "gridPos": { "h": 6, "w": 12, "x": 0, "y": 7 }, "hiddenSeries": false, "id": 310, "legend": { "alignAsTable": true, "avg": true, "current": true, "hideEmpty": true, "hideZero": true, "max": true, "min": false, "rightSide": true, "show": true, "sort": null, "sortDesc": null, "total": false, "values": true }, "lines": true, "linewidth": 1, "links": [], "nullPointMode": "null as zero", "options": { "alertThreshold": true }, "percentage": false, "pluginVersion": "7.5.11", "pointradius": 5, "points": false, "renderer": "flot", "seriesOverrides": [], "spaceLength": 10, "stack": false, "steppedLine": false, "targets": [ { "expr": "sum(rate(tidb_executor_statement_total{k8s_cluster=\"$k8s_cluster\", tidb_cluster=\"$tidb_cluster\", instance=~\"$instance\", db!=\"\"}[1m])) by (db,type)", "format": "time_series", "intervalFactor": 2, "legendFormat": "{{db}}-{{type}}", "refId": "A", "step": 30 } ], "thresholds": [], "timeFrom": null, "timeRegions": [], "timeShift": null, "title": "QPS", "tooltip": { "msResolution": false, "shared": true, "sort": 0, "value_type": "individual" }, "type": "graph", "xaxis": { "buckets": null, "mode": "time", "name": null, "show": true, "values": [] }, "yaxes": [ { "format": "short", "label": null, "logBase": 2, "max": null, "min": "0", "show": true }, { "format": "short", "label": null, "logBase": 1, "max": null, "min": null, "show": true } ], "yaxis": { "align": false, "alignLevel": null } }, { "aliasColors": {}, "bars": false, "dashLength": 10, "dashes": false, "datasource": "${DS_TEST-CLUSTER}", "decimals": null, "description": "The duration from receiving a request from the client to a database until the database executes the request and returns the result to the client.", "editable": true, "error": false, "fieldConfig": { "defaults": {}, "overrides": [] }, "fill": 1, "fillGradient": 0, "grid": {}, "gridPos": { "h": 6, "w": 12, "x": 12, "y": 7 }, "hiddenSeries": false, "id": 312, "legend": { "alignAsTable": true, "avg": true, "current": true, "hideEmpty": true, "hideZero": true, "max": true, "min": false, "rightSide": true, "show": true, "sort": null, "sortDesc": null, "total": false, "values": true }, "lines": true, "linewidth": 1, "links": [], "nullPointMode": "null as zero", "options": { "alertThreshold": true }, "percentage": false, "pluginVersion": "7.5.11", "pointradius": 5, "points": false, "renderer": "flot", "seriesOverrides": [], "spaceLength": 10, "stack": false, "steppedLine": false, "targets": [ { "expr": "sum(rate(tidb_server_handle_query_duration_seconds_sum{k8s_cluster=\"$k8s_cluster\", tidb_cluster=\"$tidb_cluster\", instance=~\"$instance\", db!=\"\"}[1m])) by (db) / sum(rate(tidb_server_handle_query_duration_seconds_count{k8s_cluster=\"$k8s_cluster\", tidb_cluster=\"$tidb_cluster\", instance=~\"$instance\", db!=\"\"}[1m])) by (db)", "format": "time_series", "intervalFactor": 2, "legendFormat": "avg-{{db}}", "refId": "A", "step": 30 } ], "thresholds": [], "timeFrom": null, "timeRegions": [], "timeShift": null, "title": "Average Query Duration Per DB", "tooltip": { "msResolution": false, "shared": true, "sort": 0, "value_type": "individual" }, "type": "graph", "xaxis": { "buckets": null, "mode": "time", "name": null, "show": true, "values": [] }, "yaxes": [ { "format": "s", "label": null, "logBase": 2, "max": null, "min": "0", "show": true }, { "format": "short", "label": null, "logBase": 1, "max": null, "min": null, "show": true } ], "yaxis": { "align": false, "alignLevel": null } }, { "aliasColors": {}, "bars": false, "dashLength": 10, "dashes": false, "datasource": "${DS_TEST-CLUSTER}", "decimals": null, "description": "The statistics of error types according to the SQL statement execution errors per second on every database.", "editable": true, "error": false, "fieldConfig": { "defaults": {}, "overrides": [] }, "fill": 1, "fillGradient": 0, "grid": {}, "gridPos": { "h": 6, "w": 12, "x": 0, "y": 7 }, "hiddenSeries": false, "id": 314, "legend": { "alignAsTable": true, "avg": true, "current": true, "hideEmpty": true, "hideZero": true, "max": true, "min": false, "rightSide": true, "show": true, "sort": null, "sortDesc": null, "total": false, "values": true }, "lines": true, "linewidth": 1, "links": [], "nullPointMode": "null as zero", "options": { "alertThreshold": true }, "percentage": false, "pluginVersion": "7.5.11", "pointradius": 5, "points": false, "renderer": "flot", "seriesOverrides": [], "spaceLength": 10, "stack": false, "steppedLine": false, "targets": [ { "expr": "sum(rate(tidb_server_execute_error_total{k8s_cluster=\"$k8s_cluster\", tidb_cluster=\"$tidb_cluster\", instance=~\"$instance\", db!=\"\"}[1m])) by (db)", "format": "time_series", "intervalFactor": 2, "legendFormat": "{{db}}", "refId": "A", "step": 30 } ], "thresholds": [], "timeFrom": null, "timeRegions": [], "timeShift": null, "title": "Failed Query Per DB", "tooltip": { "msResolution": false, "shared": true, "sort": 0, "value_type": "individual" }, "type": "graph", "xaxis": { "buckets": null, "mode": "time", "name": null, "show": true, "values": [] }, "yaxes": [ { "format": "short", "label": null, "logBase": 2, "max": null, "min": "0", "show": true }, { "format": "short", "label": null, "logBase": 1, "max": null, "min": null, "show": true } ], "yaxis": { "align": false, "alignLevel": null } } ], "repeat": null, "title": "Per DB Detail", "type": "row" },

tiup cluster reload {cluster-name} -R grafana

实现效果

手工编辑公式可查看 resource group 指标

小结

新特性 record-db-label 具有更加丰富的指标项和维度,在 resource control 场景中,可以提供细粒度的监控指标功能。

版权声明:本文内容由网络用户投稿,版权归原作者所有,本站不拥有其著作权,亦不承担相应法律责任。如果您发现本站中有涉嫌抄袭或描述失实的内容,请联系我们jiasou666@gmail.com 处理,核实后本网站将在24小时内删除侵权内容。

上一篇:TiDB数据迁移实践 DM工具使用指南
下一篇:TiDB版本7.4初体验与评测
相关文章