当前位置:网站首页>Taosi TDengine 2.6+ optimization parameters
Taosi TDengine 2.6+ optimization parameters
2022-07-30 09:20:00 【cousin code money】
taosBenchmark -I stmt -n 200000 -t 100000time dd if=/dev/vdb of=/dev/null bs=4kiostat -d -k 1 100000 Query related:
| parameters | Meaning | Default value |
|---|---|---|
| queryBufferSize | Amount of memory reserved for all concurrent query usage.Unit M | No default value |
| ratioOfQueryCores | Sets the maximum number of query threads. | The default value is 1, the default number of query threads is the same as the number of CPUs |
| ratioOfQueryThreads | Set the maximum number of query threads | 1: Indicates the maximum number of query threads equal to the number of CPU cores |
| compressColData | Query result compression transmission threshold, in bytes | -1, no compression |
Storage related:
| parameters | Meaning | Default value |
|---|---|---|
| cache | The size of the memory block, in M | 16M |
| blocks | How many cache-sized memory blocks are in each vnode (tsdb) | 6 |
| maxRows | Maximum number of records in file block | 4096 |
| maxVgroupsPerDb | Each Database can create a fixed number of vgroups, the default is the same as the number of CPU cores | |
Write related:
| parameters | Meaning | Default value |
|---|---|---|
| numOfCommitThreads | Set the maximum number of write threads | none |
Stores the INT type, which is faster than storing the Double type.Adjusted to int, indicator single table=9716 seconds, multi-table=
Stock index calculation, storage performance comparison
A. Computing server 32 cores (vCPU) 32 GiB
| Project | Data level | Number of table columns | taos service 1 | Performance Metrics | taos service 2 | Performance Metrics |
|---|---|---|---|---|---|---|
| Day K write, including read source data file | 12 million | 17 | 4 cores (vCPU) 16 GiB, PL0 200GiB (4200 IOPS) | 700 seconds | 32 cores (vCPU) 32 GiB, 200GiB (11800 IOPS) | 331 seconds |
| Indicators Write | 12 million | 94 columns, split into 8 tables | 1754 seconds, 200IPOS | 1629 seconds |
B, computing server 4 cores (vCPU) 8GiB
| Project | Data level | Number of table columns | taos service 1 | Performance Metrics | taos service 2 | Performance Metrics |
|---|---|---|---|---|---|---|
| Day K write, including read source data file | 12 million | 17 | 4 cores (vCPU) 16 GiB, PL0 200GiB (4200 IOPS) | 737 seconds | 32 cores (vCPU) 32 GiB, 200GiB (11800 IOPS) | |
| Indicators Write | 12 million | 94 columns, split into 8 tables | 5370 seconds | |||
| Indicators Write | 12 million | 94 columns, one table | 11335 seconds |
Server1
Server2

Compute Server

边栏推荐
猜你喜欢

SEER数据库中“手术变量(RX SUMM-SURG OTH REG/DIS )”下的字段解释

SRAM与DRAM的区别

MongoDB - 千万级数据脚本过滤笔记

The difference between typescript3-ts and js

typescript1-typescript是什么

The blockbuster IP that has been popular in the world for 25 years, the new work has become a script paradise

回板后,处理器不启动,怎么办?

积分专题笔记-与路径无关条件

HashSet and LinkedHashSet

Field interpretation under "Surgical variables (RX SUMM-SURG OTH REG/DIS)" in SEER database
随机推荐
JS中如何阻止事件冒泡和默认行为
SQL row-column conversion
Why does typescript2-typescript add type support to js
SQL window function
What convenience does the RFID fixed asset inventory system bring to enterprises?
一文带你玩转offer-01
typescript7-typescript常用类型
剖析SGI STL空间配置器(allocate内存分配函数)
npm指令
电源完整性的去耦和层间耦合电容
hicp第六天
电源完整性基础知识
20个电路能懂5个以上,足以证明你在电子行业混过!
opencv可以有多有趣
HashSet和LinkedHashSet
Kubernetes 在科技革命中的演变
R安装包出现error in rawtochar(block[seq_len(ns)]) :
jdbc ResultSetMetaData获取tableName问题
MongoDB - 千万级数据脚本过滤笔记
基于SSM开发实现校园疫情防控管理系统