2020-04-03 13:23:32 +00:00
|
|
|
|
---
|
2022-04-10 23:08:18 +00:00
|
|
|
|
sidebar_position: 61
|
|
|
|
|
sidebar_label: "性能测试"
|
2020-04-03 13:23:32 +00:00
|
|
|
|
---
|
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
# 性能测试 {#clickhouse-benchmark}
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
连接到ClickHouse服务器并重复发送指定的查询。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
语法:
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
|
|
|
|
``` bash
|
|
|
|
|
$ echo "single query" | clickhouse-benchmark [keys]
|
|
|
|
|
```
|
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
或
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
|
|
|
|
``` bash
|
|
|
|
|
$ clickhouse-benchmark [keys] <<< "single query"
|
|
|
|
|
```
|
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
如果要发送一组查询,请创建一个文本文件,并将每个查询的字符串放在此文件中。 例如:
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
|
|
|
|
``` sql
|
|
|
|
|
SELECT * FROM system.numbers LIMIT 10000000
|
|
|
|
|
SELECT 1
|
|
|
|
|
```
|
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
然后将此文件传递给标准输入 `clickhouse-benchmark`.
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
|
|
|
|
``` bash
|
|
|
|
|
clickhouse-benchmark [keys] < queries_file
|
|
|
|
|
```
|
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
## keys参数 {#clickhouse-benchmark-keys}
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
- `-c N`, `--concurrency=N` — Number of queries that `clickhouse-benchmark` 同时发送。 默认值:1。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
- `-d N`, `--delay=N` — Interval in seconds between intermediate reports (set 0 to disable reports). Default value: 1.
|
2020-06-13 02:48:50 +00:00
|
|
|
|
- `-h WORD`, `--host=WORD` — Server host. Default value: `localhost`. 为 [比较模式](#clickhouse-benchmark-comparison-mode) 您可以使用多个 `-h` 参数
|
2020-04-08 14:22:25 +00:00
|
|
|
|
- `-p N`, `--port=N` — Server port. Default value: 9000. For the [比较模式](#clickhouse-benchmark-comparison-mode) 您可以使用多个 `-p` 钥匙
|
2020-06-13 02:48:50 +00:00
|
|
|
|
- `-i N`, `--iterations=N` — 查询的总次数. Default value: 0.
|
|
|
|
|
- `-r`, `--randomize` — 有多个查询时,以随机顺序执行.
|
|
|
|
|
- `-s`, `--secure` — 使用TLS安全连接.
|
2020-04-08 14:22:25 +00:00
|
|
|
|
- `-t N`, `--timelimit=N` — Time limit in seconds. `clickhouse-benchmark` 达到指定的时间限制时停止发送查询。 默认值:0(禁用时间限制)。
|
|
|
|
|
- `--confidence=N` — Level of confidence for T-test. Possible values: 0 (80%), 1 (90%), 2 (95%), 3 (98%), 4 (99%), 5 (99.5%). Default value: 5. In the [比较模式](#clickhouse-benchmark-comparison-mode) `clickhouse-benchmark` 执行 [独立双样本学生的t测试](https://en.wikipedia.org/wiki/Student%27s_t-test#Independent_two-sample_t-test) 测试以确定两个分布是否与所选置信水平没有不同。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
- `--cumulative` — Printing cumulative data instead of data per interval.
|
|
|
|
|
- `--database=DATABASE_NAME` — ClickHouse database name. Default value: `default`.
|
2020-04-08 14:22:25 +00:00
|
|
|
|
- `--json=FILEPATH` — JSON output. When the key is set, `clickhouse-benchmark` 将报告输出到指定的JSON文件。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
- `--user=USERNAME` — ClickHouse user name. Default value: `default`.
|
|
|
|
|
- `--password=PSWD` — ClickHouse user password. Default value: empty string.
|
2020-04-08 14:22:25 +00:00
|
|
|
|
- `--stacktrace` — Stack traces output. When the key is set, `clickhouse-bencmark` 输出异常的堆栈跟踪。
|
2020-06-13 02:48:50 +00:00
|
|
|
|
- `--stage=WORD` — 查询请求的服务端处理状态. 在特定阶段Clickhouse会停止查询处理,并返回结果给`clickhouse-benchmark`。 可能的值: `complete`, `fetch_columns`, `with_mergeable_state`. 默认值: `complete`.
|
2020-04-03 13:23:32 +00:00
|
|
|
|
- `--help` — Shows the help message.
|
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
如果你想在查询时应用上述的部分参数 [设置](../../operations/settings/index.md) ,请将它们作为键传递 `--<session setting name>= SETTING_VALUE`. 例如, `--max_memory_usage=1048576`.
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
## 输出 {#clickhouse-benchmark-output}
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
默认情况下, `clickhouse-benchmark` 按照 `--delay` 参数间隔输出结果。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
报告示例:
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
|
|
|
|
``` text
|
|
|
|
|
Queries executed: 10.
|
|
|
|
|
|
|
|
|
|
localhost:9000, queries 10, QPS: 6.772, RPS: 67904487.440, MiB/s: 518.070, result RPS: 67721584.984, result MiB/s: 516.675.
|
|
|
|
|
|
|
|
|
|
0.000% 0.145 sec.
|
|
|
|
|
10.000% 0.146 sec.
|
|
|
|
|
20.000% 0.146 sec.
|
|
|
|
|
30.000% 0.146 sec.
|
|
|
|
|
40.000% 0.147 sec.
|
|
|
|
|
50.000% 0.148 sec.
|
|
|
|
|
60.000% 0.148 sec.
|
|
|
|
|
70.000% 0.148 sec.
|
|
|
|
|
80.000% 0.149 sec.
|
|
|
|
|
90.000% 0.150 sec.
|
|
|
|
|
95.000% 0.150 sec.
|
|
|
|
|
99.000% 0.150 sec.
|
|
|
|
|
99.900% 0.150 sec.
|
|
|
|
|
99.990% 0.150 sec.
|
|
|
|
|
```
|
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
在结果报告中,您可以找到:
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
- 查询数量:参见`Queries executed:`字段。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
- 状态码(按顺序给出):
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
- ClickHouse服务器的连接信息。
|
2020-04-08 14:22:25 +00:00
|
|
|
|
- 已处理的查询数。
|
2021-07-29 15:20:55 +00:00
|
|
|
|
- QPS:服务端每秒处理的查询数量
|
|
|
|
|
- RPS:服务器每秒读取多少行
|
|
|
|
|
- MiB/s:服务器每秒读取多少字节的数据
|
|
|
|
|
- 结果RPS:服务端每秒生成多少行的结果集数据
|
2021-07-29 15:27:50 +00:00
|
|
|
|
- 结果MiB/s.服务端每秒生成多少字节的结果集数据
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
- 查询执行时间的百分比。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
## 对比模式 {#clickhouse-benchmark-comparison-mode}
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
`clickhouse-benchmark` 可以比较两个正在运行的ClickHouse服务器的性能。
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-06-13 02:48:50 +00:00
|
|
|
|
要使用对比模式,分别为每个服务器配置各自的`--host`, `--port`参数。`clickhouse-benchmark` 会根据设置的参数建立到各个Server的连接并发送请求。每个查询请求会随机发送到某个服务器。输出结果会按服务器分组输出
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
2020-04-08 14:22:25 +00:00
|
|
|
|
## 示例 {#clickhouse-benchmark-example}
|
2020-04-03 13:23:32 +00:00
|
|
|
|
|
|
|
|
|
``` bash
|
|
|
|
|
$ echo "SELECT * FROM system.numbers LIMIT 10000000 OFFSET 10000000" | clickhouse-benchmark -i 10
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
``` text
|
|
|
|
|
Loaded 1 queries.
|
|
|
|
|
|
|
|
|
|
Queries executed: 6.
|
|
|
|
|
|
|
|
|
|
localhost:9000, queries 6, QPS: 6.153, RPS: 123398340.957, MiB/s: 941.455, result RPS: 61532982.200, result MiB/s: 469.459.
|
|
|
|
|
|
|
|
|
|
0.000% 0.159 sec.
|
|
|
|
|
10.000% 0.159 sec.
|
|
|
|
|
20.000% 0.159 sec.
|
|
|
|
|
30.000% 0.160 sec.
|
|
|
|
|
40.000% 0.160 sec.
|
|
|
|
|
50.000% 0.162 sec.
|
|
|
|
|
60.000% 0.164 sec.
|
|
|
|
|
70.000% 0.165 sec.
|
|
|
|
|
80.000% 0.166 sec.
|
|
|
|
|
90.000% 0.166 sec.
|
|
|
|
|
95.000% 0.167 sec.
|
|
|
|
|
99.000% 0.167 sec.
|
|
|
|
|
99.900% 0.167 sec.
|
|
|
|
|
99.990% 0.167 sec.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Queries executed: 10.
|
|
|
|
|
|
|
|
|
|
localhost:9000, queries 10, QPS: 6.082, RPS: 121959604.568, MiB/s: 930.478, result RPS: 60815551.642, result MiB/s: 463.986.
|
|
|
|
|
|
|
|
|
|
0.000% 0.159 sec.
|
|
|
|
|
10.000% 0.159 sec.
|
|
|
|
|
20.000% 0.160 sec.
|
|
|
|
|
30.000% 0.163 sec.
|
|
|
|
|
40.000% 0.164 sec.
|
|
|
|
|
50.000% 0.165 sec.
|
|
|
|
|
60.000% 0.166 sec.
|
|
|
|
|
70.000% 0.166 sec.
|
|
|
|
|
80.000% 0.167 sec.
|
|
|
|
|
90.000% 0.167 sec.
|
|
|
|
|
95.000% 0.170 sec.
|
|
|
|
|
99.000% 0.172 sec.
|
|
|
|
|
99.900% 0.172 sec.
|
|
|
|
|
99.990% 0.172 sec.
|
|
|
|
|
```
|