TDengine
TDengine 是一款高性能、分布式、支持 SQL 的时序数据库 (Database)。在开通采集器之前请先熟悉 TDengine 基本概念
TDengine 采集器需要的连接 taos_adapter 才可以正常工作，taosAdapter 从 TDengine v2.4.0.0 版本开始成为 TDengine 服务端软件 的一部分，本文主要是指标集的详细介绍。
配置¶
采集器配置¶
进入 DataKit 安装目录下的 conf.d/samples 目录，复制 tdengine.conf.sample 并命名为 tdengine.conf。示例如下：
[[inputs.tdengine]]
## adapter restApi Addr, example: http://taosadapter.test.com (Required)
adapter_endpoint = "http://<FQND>:6041"
user = "<userName>"
password = "<pw>"
## log_files: TdEngine log file path or dirName (optional).
## log_files = ["tdengine_log_path.log"]
## pipeline = "tdengine.p"
## Set true to enable election
election = true
## add tag (optional)
[inputs.tdengine.tags]
## Different clusters can be distinguished by tag. Such as testing,product,local ,default is 'testing'
# cluster_name = "testing"
# some_tag = "some_value"
# more_tag = "some_other_value"
配置好后，重启 DataKit 即可。
目前可以通过 ConfigMap 方式注入采集器配置来开启采集器。
Info
- 连接 taoAdapter 之前请先确定端口是开放的。并且连接用户需要有 read 权限。
- 若仍连接失败，请参考此处。
指标¶
tdengine¶
TDengine cluster, database, data node, request, and taosAdapter metrics collected from system log tables and show commands.
| Tags & Fields | Description |
|---|---|
| client_ip ( tag) |
Client IP |
| cluster_name ( tag) |
Cluster name |
| database_name ( tag) |
Database name |
| dnode_ep ( tag) |
Data node name, generally equivalent to end_point |
| end_point ( tag) |
Remote address name, the general naming rule is (host:port) |
| endpoint ( tag) |
taosAdapter endpoint |
| first_ep ( tag) |
First endpoint |
| host ( tag) |
Host name |
| offline_reason ( tag) |
Reason the data node is offline |
| role ( tag) |
Management node role |
| role_time ( tag) |
Management node role start time |
| status ( tag) |
Data node status |
| status_code ( tag) |
HTTP response status code |
| version ( tag) |
Version |
| vgroup_id ( tag) |
VGroup ID |
| client_ip_count | Client IP request statistics Type: float | (gauge) Unit: count Tagged by: client_ip |
| cpu_cores | Total number of CPU cores per data node Type: float | (gauge) Unit: count Tagged by: dnode_ep |
| cpu_engine | CPU usage per data node Type: float | (gauge) Unit: percent,percent Tagged by: dnode_ep |
| cpu_percent | Adapter occupies CPU usage Type: float | (gauge) Unit: percent,percent Tagged by: endpoint |
| cpu_system | CPU system usage of data nodes Type: float | (gauge) Unit: count Tagged by: dnode_ep |
| database_count | Total number of databases Type: float | (gauge) Unit: count |
| disk_percent | Data node disk usage percentage Type: float | (gauge) Unit: percent,percent Tagged by: dnode_ep |
| disk_total | Total disk size of data nodes Type: float | (gauge) Unit: digital,GB Tagged by: dnode_ep |
| disk_used | Disk usage of data nodes Type: float | (gauge) Unit: digital,GB Tagged by: dnode_ep |
| dnodes_alive | Total number of dnodes in ready state Type: float | (gauge) Unit: count |
| dnodes_total | Total number of dnodes(data nodes) in cluster Type: float | (gauge) Unit: count |
| expire_time | Time until grants expire in seconds Type: int | (gauge) Unit: time,s |
| id | TDengine node ID Type: float | (gauge) Unit: N/A Tagged by: end_point, offline_reason, role, role_time, status |
| io_read_taosd | Average data size of IO reads per second Type: float | (gauge) Unit: digital,MB Tagged by: dnode_ep |
| io_write_taosd | Average data size of IO writes per second Type: float | (gauge) Unit: digital,MB Tagged by: dnode_ep |
| master_uptime | Seconds of master's uptime Type: float | (gauge) Unit: time,s Tagged by: first_ep, version |
| mem_engine | Memory usage of tdengine Type: float | (gauge) Unit: digital,MB Tagged by: dnode_ep |
| mem_engine_percent | taosd memory usage percentageType: float | (gauge) Unit: percent,percent Tagged by: dnode_ep |
| mem_percent | Adapter memory usage Type: float | (gauge) Unit: percent,percent Tagged by: endpoint |
| mem_system | Available memory on the server Type: float | (gauge) Unit: digital,MB Tagged by: dnode_ep |
| mem_total | Total memory of server Type: float | (gauge) Unit: digital,GB Tagged by: dnode_ep |
| mnodes_alive | Total number of mnodes in ready state Type: float | (gauge) Unit: count |
| mnodes_total | Total number of mnodes(management nodes) in cluster Type: float | (gauge) Unit: count |
| net_in | IO rate of the ingress network Type: float | (gauge) Unit: digital,KB Tagged by: dnode_ep |
| net_out | IO rate of egress network Type: float | (gauge) Unit: digital,KB Tagged by: dnode_ep |
| req_http | Total number of requests via HTTP Type: float | (gauge) Unit: count |
| req_http_rate | HTTP request rate Type: float | (gauge) Unit: count |
| req_insert_batch_rate | Number of batch insertions divided by monitor interval Type: float | (gauge) Unit: count Tagged by: dnode_ep |
| req_insert_rate | Number of insert queries received per dnode divided by monitor interval Type: float | (gauge) Unit: count Tagged by: dnode_ep |
| req_select | Number of select queries received per dnode Type: float | (gauge) Unit: count Tagged by: dnode_ep |
| req_select_rate | Number of select queries received per dnode divided by monitor interval Type: float | (gauge) Unit: count Tagged by: dnode_ep |
| request_in_flight | Number of requests being sorted Type: float | (gauge) Unit: count |
| status_code | Status code returned by the request Type: float | (gauge) Unit: count |
| table_count | Total number of tables in the database Type: float | (gauge) Unit: count |
| tables_count | Number of tables per vgroup Type: float | (gauge) Unit: count Tagged by: database_name |
| timeseries_total | Total time series Type: float | (gauge) Unit: count |
| timeseries_used | Time series used Type: float | (gauge) Unit: count |
| total_req_count | Total adapter requests Type: float | (gauge) Unit: count Tagged by: endpoint, status_code |
| uptime | Data node uptime in seconds Type: float | (gauge) Unit: time,s Tagged by: dnode_ep |
| vgroups_alive | Total number of vgroups in ready state Type: float | (gauge) Unit: count |
| vgroups_total | Total number of vgroups in cluster Type: float | (gauge) Unit: count |
| vnodes | The number of virtual node groups contained in a single data node Type: float | (gauge) Unit: count Tagged by: end_point, offline_reason, status |
| vnodes_alive | Total number of vnode in ready state Type: float | (gauge) Unit: count |
| vnodes_num | Total number of virtual nodes per data node Type: float | (gauge) Unit: count Tagged by: dnode_ep |
| vnodes_total | Total number of vnode in cluster Type: float | (gauge) Unit: count |
- 数据库中有些表中没有
ts字段，DataKit 会使用当前采集的时间。