搜索到
27
篇与
的结果
-
搭建MySQL高可用集群 MySQL InnoDB Cluster 搭建过程基于Ubuntu 20.04系统搭建系统防火墙均已关闭为了方便管理，主节点和从节点root密码统一使用一个模式：单主多从graph LR A[Application/Client] --> B[Mysql Router] B --> A B --> C[MySQL Master R/W] C --> |Group Replication| D[Slave01 R/O] C --> |Group Replication| E[Slave02 R/O] C --> |Group Replication| F[Slave02 R/O] C --> |Group Replication| G[Slave02 R/O] 节点IP作用node01192.168.121.135主node02192.168.121.136从node03192.168.121.137从node04192.168.121.138路由管理系统配置(所有服务器)在每台服务器上配置hosts，编辑/etc/hosts文件，加入以下内容192.168.121.135 node01 192.168.121.136 node02 192.168.121.137 node03 192.168.121.138 node04软件安装主节点、从节点、路由管理节点初始化，安装MySQL官方apt源sudo wget https://repo.mysql.com/mysql-apt-config_0.8.19-1_all.deb && sudo dpkg -i mysql-apt-config_0.8.19-1_all.deb && sudo apt-get update主节点、从节点安装mysql-server、mysql-shellsudo apt install mysql-server mysql-shell -y路由管理节点安装mysql-shell、mysql-routersudo apt install mysql-router mysql-shell -yMySQL配置(主、从服务器)修改MySQL配置文件/etc/mysql/mysql.conf.d/mysqld.cnf(主节点)[mysqld] pid-file = /var/run/mysqld/mysqld.pid socket = /var/run/mysqld/mysqld.sock datadir = /var/lib/mysql log-error = /var/log/mysql/error.log default-storage-engine = INNODB character-set-server = utf8mb4 port = 3306 #复制框架 server_id = 1 gtid_mode = ON enforce_gtid_consistency = ON master_info_repository = TABLE relay_log_info_repository = TABLE binlog_checksum = NONE log_slave_updates = ON log_bin = binlog binlog_format = ROW max_connections = 100 max_allowed_packet = 100M #组复制设置 #server必须为每个事务收集写集合，并使用XXHASH64哈希算法将其编码为散列 transaction_write_set_extraction = XXHASH64 #告知插件加入或创建组命名，UUID loose-group_replication_group_name = "aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa" #server启动时不自启组复制,为了避免每次启动自动引导具有相同名称的第二个组,所以设置为OFF。 loose-group_replication_start_on_boot = off #告诉插件使用IP地址，端口33061用于接收组中其他成员转入连接 loose-group_replication_local_address = "node01:33061" #启动组server，种子server，加入组应该连接这些的ip和端口；其他server要加入组得由组成员同意 loose-group_replication_group_seeds = "node01:33061,node02:33061,node03:33061" loose-group_replication_ip_whitelist = "node01,node02,node03,node04" loose-group_replication_bootstrap_group = off # 使用MGR的单主模式 loose-group_replication_single_primary_mode = on loose-group_replication_enforce_update_everywhere_checks = off disabled_storage_engines = MyISAM,BLACKHOLE,FEDERATED,CSV,ARCHIVE从节点也需要修改，只需要把配置文件中的server_id、loose-group_replication_local_address修改为对应的值例如node02server_id = 2 loose-group_replication_local_address = "node02:33061"以此类推重启MySQLsudo service mysql restart使root用户支持远程登录update mysql.user set host='%' where user='root';赋予管理集群所需的权限GRANT ALL ON *.* TO 'root'@'%' WITH GRANT OPTION;刷新权限flush privileges;配置本地实例在所有节点服务器上配置本地实例，使用mysqlsh工具dba.configureLocalInstance('root@localhost:3306', {'password': 'root', 'interactive': false});检查实例状态dba.checkInstanceConfiguration("root@localhost:3306");如果输出以下内容说明配置成功。The instance 'node01:3306' is valid to be used in an InnoDB cluster. { "status": "ok" }创建集群在主节点上创建集群，同样使用mysqlsh工具# 连接实例 shell.connect('root@localhost:3306'); # 创建一个集群，命名为 'myCluster' var cluster = dba.createCluster('myCluster'); # 查看集群状态 cluster.status();输出以下内容就表示创建成功{ "clusterName": "myCluster", "defaultReplicaSet": { "name": "default", "primary": "node01:3306", "ssl": "REQUIRED", "status": "OK_NO_TOLERANCE", "statusText": "Cluster is NOT tolerant to any failures.", "topology": { "node01:3306": { "address": "node01:3306", "memberRole": "PRIMARY", "mode": "R/W", "readReplicas": {}, "replicationLag": null, "role": "HA", "status": "ONLINE", "version": "8.0.27" } }, "topologyMode": "Single-Primary" }, "groupInformationSourceMember": "node01:3306" }从节点加入集群在主节点中打开mysqlsh工具# 连接实例 shell.connect('root@localhost:3306'); # 根据集群名字获取集群对象 var cluster = dba.getCluster('myCluster'); # 将node02节点加入集群 cluster.addInstance('root@node02:3306');如果出现WARNING: A GTID set check of the MySQL instance at 'node02:3306' determined that it contains transactions that do not originate from the cluster, which must be discarded before it can join the cluster. node02:3306 has the following errant GTIDs that do not exist in the cluster: bbdb0716-32e7-11ec-86e0-000c29040a05:1-2 WARNING: Discarding these extra GTID events can either be done manually or by completely overwriting the state of node02:3306 with a physical snapshot from an existing cluster member. To use this method by default, set the 'recoveryMethod' option to 'clone'. Having extra GTID events is not expected, and it is recommended to investigate this further and ensure that the data can be removed prior to choosing the clone recovery method. Please select a recovery method [C]lone/[A]bort (default Abort):输入C继续，之后提示出下面内容说明加入成功The instance 'node02:3306' was successfully added to the cluster.其他从节点加入的方式一致查询集群状态cluster.status()输出下面内容{ "clusterName": "myCluster", "defaultReplicaSet": { "name": "default", "primary": "node01:3306", "ssl": "REQUIRED", "status": "OK", "statusText": "Cluster is ONLINE and can tolerate up to ONE failure.", "topology": { "node01:3306": { "address": "node01:3306", "memberRole": "PRIMARY", "mode": "R/W", "readReplicas": {}, "replicationLag": null, "role": "HA", "status": "ONLINE", "version": "8.0.27" }, "node02:3306": { "address": "node02:3306", "memberRole": "SECONDARY", "mode": "R/O", "readReplicas": {}, "replicationLag": null, "role": "HA", "status": "ONLINE", "version": "8.0.27" }, "node03:3306": { "address": "node03:3306", "memberRole": "SECONDARY", "mode": "R/O", "readReplicas": {}, "replicationLag": null, "role": "HA", "status": "ONLINE", "version": "8.0.27" } }, "topologyMode": "Single-Primary" }, "groupInformationSourceMember": "node01:3306" }可以看到，所有节点均已加入集群，并且正常运行配置 MySQL Router创建配置文件保存目录sudo mkdir /mysqlrouter将创建的目录授权给mysqlrouter用户编辑/etc/apparmor.d/usr.bin.mysqlrouter文件，加入以下内容# Allow config file access /mysqlrouter/ rw, /mysqlrouter/** rw,重启apparmorsudo /etc/init.d/apparmor restart配置 MySQL Router，生成配置文件到/data/mysqlrouter目录sudo mysqlrouter --bootstrap node01:3306 --directory /mysqlrouter --user=root --conf-use-sockets --force--conf-use-sockets 表示使用 unix sockets 通信--directory 指定配置文件保存目录输出下面内容表示创建成功# Bootstrapping MySQL Router instance at '/mysqlrouter'... - Creating account(s) (only those that are needed, if any) - Verifying account (using it to run SQL queries that would be run by Router) - Storing account in keyring - Adjusting permissions of generated files - Creating configuration /mysqlrouter/mysqlrouter.conf # MySQL Router configured for the InnoDB Cluster 'myCluster' After this MySQL Router has been started with the generated configuration $ mysqlrouter -c /mysqlrouter/mysqlrouter.conf InnoDB Cluster 'myCluster' can be reached by connecting to: ## MySQL Classic protocol - Read/Write Connections: localhost:6446, /mysqlrouter/mysql.sock - Read/Only Connections: localhost:6447, /mysqlrouter/mysqlro.sock ## MySQL X protocol - Read/Write Connections: localhost:6448, /mysqlrouter/mysqlx.sock - Read/Only Connections: localhost:6449, /mysqlrouter/mysqlxro.sock启动 MySQL Routersudo /mysqlrouter/start.sh启动完成后会监听下面几个端口端口socket文件协议服务6446/mysqlrouter/mysql.sockMySQLMySQL 读写6447/mysqlrouter/mysqlro.sockMySQLMySQL 只读6448/mysqlrouter/mysqlx.sockMySQL XMySQL X 读写6449/mysqlrouter/mysqlxro.sockMySQL XMySQL X 只读8443无HTTPSWeb 管理验证集群登录后，新建一个数据库和表，往里面写进数据，查看从节点数据会不会同步；关闭node01的mysql服务，route将主节点自动切换到node02，node02从只读变为可读写，重新启动node01 mysql后，node01变为只读模式。 -
[2021]PHP循环性能比较 导语时间飞速流逝，转眼已经过去了3年，这几年科技在快速发展，PHP也来到了8.0.10版本。回望这几年，大数据时代悄然降临，面对海量的数据，我们迫切的需要更为快速的运算性能；基础硬件的升级、软件算法的迭代，这些都是必不可少的部分。编程语言也在不断进化着，但是我们写的代码真的高效吗？你真的会用这门语言吗？带着这些疑问来到我们今天的测试项目。循环语句是我们在编程中经常书写的代码，我们通常会使用循环对数组、集合、列表进行遍历，但是每种循环语句的实现各不相同，今天我们就对for、foreach这两个最常用的语句进行一个简单的性能测试。这次测试不包含内置函数的测试，例如array_map等回调类型的函数，也不对while、do...while语句进行测试，毕竟用while、do...while遍历数据的情况较少。其实早在2018年的时候做过这样的一个测试，但是不严谨，具体可以看之前这篇文章： PHP循环性能测试测试方法使用for、foreach、foreach-key三种方式进行嵌套空循环测试，每一层循环1000次，且每种方式运行5次，记录每次运行的耗时，取最小值作为最终测试结果。测试代码算法例子：$test_result = []; for($t = 0; $t < 5; $t++) { $start_time = microtime(true); for($i = 0; $i < 1000; $i++) { for($j = 0; $j < 1000; $j++) { // do something } } $end_time = microtime(true); $test_result[] = ($end_time-$start_time)*1000; } echo 'for:'.min($test_result).PHP_EOL;测试环境硬件环境 CPU单核 Intel(R) Xeon(R) Platinum 8255C CPU @ 2.50GHz内存2GB硬盘60GB SSD云硬盘操作系统Ubuntu Server 20.04 LTS 64bit软件环境PHP 5.6.40PHP 7.4.23PHP 8.0.10完整测试代码<?php function _for($test_count = 5, $loop_count = 1000) { $test_result = []; for($t = 0; $t < $test_count; $t++) { $start_time = microtime(true); for($i = 0; $i < $loop_count; $i++) { for($j = 0; $j < $loop_count; $j++) { // do something } } $end_time = microtime(true); $test_result[] = ($end_time-$start_time)*1000; } echo 'for:'.min($test_result).PHP_EOL; } function _foreach($test_count = 5, $loop_count = 1000) { $test_result = []; $array = range(1, $loop_count); for($t = 0; $t < $test_count; $t++) { $start_time = microtime(true); foreach($array as $value) { foreach($array as $value2) { // do something } } $end_time = microtime(true); $test_result[] = ($end_time-$start_time)*1000; } echo 'foreach:'.min($test_result).PHP_EOL; } function _foreach_key($test_count = 5, $loop_count = 1000) { $test_result = []; $array = range(1, $loop_count); for($t = 0; $t < $test_count; $t++) { $start_time = microtime(true); foreach($array as $key => $value) { foreach($array as $key2 => $value2) { // do something } } $end_time = microtime(true); $test_result[] = ($end_time-$start_time)*1000; } echo 'foreach-key:'.min($test_result).PHP_EOL; } _for(); _foreach(); _foreach_key();测试结果PHP5.6.40PHP7.4.23PHP8.0.10测试方法PHP 5.6.40PHP 7.4.23PHP 8.0.10for21.63696289062513.2529735565194.356861114502foreach46.97203636169419.380092620856.3300132751465foreach-key53.55119705200234.90209579467810.4660987854以上耗时结果均为毫秒单位，值越小越好从上面的结果可以看出，新版本对性能的提升确实非常大，基本上都是质的提升。如果可以尽量使用最新的稳定版。PHP8对比PHP7基本是3倍的性能提升。结束语这次测试结果还是让我很意外，本以为PHP7性能提升得已经够多了，没想到PHP8直接碾压，再加上PHP8引入的JIT(Just In Time)，让PHP8非常乐观，不过到目前我还没弄懂在哪些情况下才会用到JIT，如果后面有时间我会专门写一篇文章来测试JIT。这篇文章的目的只是参考作用，毕竟只是空循环，实际应用中可能会复杂很多。但有一点可以肯定，新版本带来的性能提升是毋庸置疑的，大家可以放心食用~
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
![[2021]PHP循环性能比较](https://static.jindashen.cn/usr/uploads/2021/09/2587412554.png)
![[免费]JetBrains全系列软件激活教程激活码](https://unpkg.com/typecho-joe-next@6.0.0/assets/thumb/6.jpg)
![[2018]PHP循环性能比较](https://unpkg.com/typecho-joe-next@6.0.0/assets/thumb/20.jpg)
