当前位置:网站首页>Matrixdb V4.5.0 was launched with a new mars2 storage engine!
Matrixdb V4.5.0 was launched with a new mars2 storage engine!
2022-07-05 06:06:00 【Ymatrix official technology community】
After nearly 4 Months of hard work , all-new MatrixDB v4.5.0 Official release .
this MatrixDB v4.5.0 Welcome many major updates , Include : Launch a new storage engine MARS2, Add query information statistics component 、 New support sliding window And MatrixDB on Kubernetes Deployment and other functional features ,Mxbench 、MatrixGate Upgrade both , And fixed some problems encountered in production practice .MatrixDB In terms of functionality and stability , Continuous and stable improvement .
01 New launch MARS2 Storage engine
MARS2 Storage engine in MARS1 Develop on the basis of , It's solved MARS1 The problem of guiding tables . In order to achieve the effect of dynamic sorting ,MARS2 The internal of storage draws lessons from LSM TREE The concept of , Implemented by a two-tier class LSM structure . Classes at each level LSM The structure consists of two sorting units ( Logical tape set or tapesets) form , Each sorting unit consists of several logical tapes (tape) form . Finally, dynamic sorting is realized , The effect of dynamic compression .
02 Add query statistics mxstat
Add query statistics component (mxstat), By aggregating the same queries , To count query execution times 、 Time consuming 、 Resource occupancy information, etc . Through this function , Users can perform on the database SQL Statistics and monitoring , And better understand the resource consumption of various queries , Then we can locate and analyze slow queries .
03 Support sliding window flow calculation
Sliding window is a common function of flow computing scene . In the scenario of stream computing data , Through the sliding window, the data in the recent period of time can be continuously analyzed and calculated . Unlike scrolling windows, non overlapping , Time windows of sliding windows can be overlapped . Sliding windows are usually used in conjunction with monitoring alarms , When the data in the latest period of time meets the preset conditions , The data server will send an alarm message to the client .
MatrixDB v4.5.0 Add sliding window flow calculation , It can realize real-time analysis and monitoring of convection data , And support the subscription and publication of monitoring data , When the data is abnormal, you can get the alarm message through subscription .
04 Mxbench Optimize upgrade , Support flexible configuration
Mxbench yes MatrixDB Pressure test tool for data loading and query , According to the number of devices given by the user 、 Time range 、 The indicator quantity configuration can quickly generate random data , Automatically create data tables , Load and query data serially or concurrently .
this Mxbench It's upgraded , Support users to flexibly configure indicator types 、 Acquisition frequency 、 Null rate 、 Randomness, etc , Users can also specify the number of threads to query 、 Custom query statements, etc .Mxbench You can run... From the command line , You can also run through the configuration file .
05 MatrixGate Upgrade again
MatrixGate abbreviation mxgate, yes MatrixDB Built in high-performance data loading tool . This upgrade ,mxgate newly added :
Support for CSV Any column of the file , perhaps JSON Any of the files Value, Perform timestamp type conversion (timestamp), The conversion accuracy supports second level , millisecond , Microsecond and nanosecond ;
newly added Watch function , Can be in mxgate In operation , see mxgate Real time indicators and historical indicators of data loading , The data loading performance is clear at a glance , At the same time, it also greatly reduces the difficulty of troubleshooting problems related to data loading ;
Upsert Support boolean data type .
06 Support MatrixDB on K8s Deploy
MatrixDB v4.5.0 New support through Operator Pattern , Realize in Kubernetes Rapid deployment on a cluster MatrixDB The function of . Users can use the Kubernetes Install in cluster matrixdb-operator The way , Implement multiple MatrixDB Rapid deployment of clusters .
More information
You can use the official website (ymatrix.cn) Download and get the free community version , Or contact us for commercial version support . If you want to know more about MatrixDB v4.5.0 You can refer to MatrixDB file To deploy the experience .
Download address : The world's first super integrated time series database Or check out MatrixDB file
About MatrixDB
MatrixDB It's a lightweight 、 High performance 、 One stop hyperfusion time series database , Through Microkernel (microdb) The technical architecture realizes one library and multiple uses , It is provided on a set of database architecture , For multimodal data 、 Multiple scene types 、 Unified support for multiple complex queries , Help enterprises significantly reduce the procurement of data infrastructure 、 Threshold of use and operation and maintenance . at present ,MatrixDB It has been successfully applied to intelligent connected vehicles 、 Intelligent manufacturing 、 Industrial Internet 、 Smart energy 、 Smart city and other scenes , And won a lot of wealth 500 Strong enterprise recognition .
边栏推荐
- LeetCode 0108.将有序数组转换为二叉搜索树 - 数组中值为根,中值左右分别为左右子树
- leetcode-1200:最小绝对差
- Sword finger offer 05 Replace spaces
- Introduction et expérience de wazuh open source host Security Solution
- [rust notes] 14 set (Part 1)
- leetcode-556:下一个更大元素 III
- Implement a fixed capacity stack
- CCPC Weihai 2021m eight hundred and ten thousand nine hundred and seventy-five
- Simple knapsack, queue and stack with deque
- Sword finger offer 05 Replace spaces
猜你喜欢
做 SQL 性能优化真是让人干瞪眼
Dichotomy, discretization, etc
2017 USP Try-outs C. Coprimes
Sword finger offer 53 - ii Missing numbers from 0 to n-1
leetcode-6108:解密消息
Scope of inline symbol
Sword finger offer 04 Search in two-dimensional array
Appium基础 — 使用Appium的第一个Demo
Sword finger offer 05 Replace spaces
【Jailhouse 文章】Jailhouse Hypervisor
随机推荐
Common optimization methods
Daily question - Search two-dimensional matrix PS two-dimensional array search
CF1637E Best Pair
AtCoder Grand Contest 013 E - Placing Squares
leetcode-31:下一个排列
leetcode-9:回文数
1.14 - 流水线
Typical use cases for knapsacks, queues, and stacks
Sword finger offer 09 Implementing queues with two stacks
【云原生】微服务之Feign自定义配置的记录
6. Logistic model
[jailhouse article] performance measurements for hypervisors on embedded ARM processors
SQLMAP使用教程(一)
Collection: programming related websites and books
传统数据库逐渐“难适应”,云原生数据库脱颖而出
shared_ Repeated release heap object of PTR hidden danger
CF1634E Fair Share
Introduction et expérience de wazuh open source host Security Solution
【Rust 笔记】17-并发(上)
Over fitting and regularization