前言
本文主要是講解EF Core3.0+ 通過攔截器實現讀寫分離與SQL日誌記錄
注意攔截器只有EF Core3.0+ 支援,2.1請考慮上下文工廠的形式實現.
說點題外話..
一晃又大半年沒更新技術部落格..唉,去年一年發生了太多事情..博主真的 一言難盡..
有興趣的可以去看看:記錄一下,也許是轉折,也許是結束,也許是新希望的一年
正文
1.通過攔截器實現讀寫分離
先講一下本文實現的方式吧
SQL 通過資料庫本身的功能 實現主從備份 大概原理如圖:
EF Core在查詢的時候通過DbCommandInterceptor 攔截器(PS:這個功能在EF6.0+中也實現了)來攔截對資料庫的訪問,從而切換主從資料庫
下面直接上程式碼吧
首先我們建立一個類 繼承DbCommandInterceptor:
public class DbMasterSlaveCommandInterceptor : DbCommandInterceptor { private string _masterConnectionString; private string _slaveConnectionString; public DbMasterSlaveCommandInterceptor(string masterConnectionString, string slaveConnectionString) { _masterConnectionString = masterConnectionString; _slaveConnectionString = slaveConnectionString; } }
通過建構函式傳遞主庫連線地址與從庫地址(可有多個 通過"|"分割)
新增一個隨機分配從表讀取連線的方法(PS:這裡只是demo所以很簡陋的隨機,如果正式要用,應包含權重判斷,定時心跳從庫連線情況,請自行修改):
/// <summary> /// 通過隨機數分配獲取多個從庫 /// </summary> /// <returns></returns> private string GetSlaveConnectionString() { var readArr = _slaveConnectionString.Split(new char[] { '|' }, StringSplitOptions.RemoveEmptyEntries); var resultConn = string.Empty; if (readArr != null && readArr.Any()) { resultConn = readArr[Convert.ToInt32(Math.Floor((double)new Random().Next(0, readArr.Length)))]; } return resultConn; }
新增判斷是否主從操作連線方法:
private void UpdateToSlave(DbCommand command) { //判斷是否配置了主從分離 if (!string.IsNullOrWhiteSpace(GetSlaveConnectionString()))//如果配置了讀寫分離,就進入判斷 { //判斷是否為插入語句(EF 插入語句會通過Reader執行並查詢主鍵),否則進入 if (command.CommandText.ToLower().StartsWith("insert", StringComparison.InvariantCultureIgnoreCase) == false) { // 判斷當前會話是否處於分散式事務中 bool isDistributedTran = Transaction.Current != null && Transaction.Current.TransactionInformation.Status != TransactionStatus.Committed; //判斷該 context 是否處於普通資料庫事務中 bool isDbTran = command.Transaction != null; //如果不處於事務中,則執行從伺服器查詢 if (!isDbTran && !isDistributedTran) { command.Connection.Close(); command.Connection.ConnectionString = GetSlaveConnectionString(); command.Connection.Open(); } } } }
過載DbCommandInterceptor當中的攔截方法,程式碼如下:
//如果是寫入,則正常執行 public override InterceptionResult<int> NonQueryExecuting(DbCommand command, CommandEventData eventData, InterceptionResult<int> result) { return base.NonQueryExecuting(command, eventData, result); } public override Task<InterceptionResult<int>> NonQueryExecutingAsync(DbCommand command, CommandEventData eventData, InterceptionResult<int> result, CancellationToken cancellationToken = default) { return base.NonQueryExecutingAsync(command, eventData, result, cancellationToken); } public override InterceptionResult<DbDataReader> ReaderExecuting(DbCommand command, CommandEventData eventData, InterceptionResult<DbDataReader> result) { this.UpdateToSlave(command); return base.ReaderExecuting(command, eventData, result); } public override Task<InterceptionResult<DbDataReader>> ReaderExecutingAsync(DbCommand command, CommandEventData eventData, InterceptionResult<DbDataReader> result, CancellationToken cancellationToken = default) { this.UpdateToSlave(command); return base.ReaderExecutingAsync(command, eventData, result, cancellationToken); } public override InterceptionResult<object> ScalarExecuting(DbCommand command, CommandEventData eventData, InterceptionResult<object> result) { this.UpdateToSlave(command); return base.ScalarExecuting(command, eventData, result); } public override Task<InterceptionResult<object>> ScalarExecutingAsync(DbCommand command, CommandEventData eventData, InterceptionResult<object> result, CancellationToken cancellationToken = default) { this.UpdateToSlave(command); return base.ScalarExecutingAsync(command, eventData, result, cancellationToken); }
最後在EF core的上下文中注入攔截器(PS:我這裡使用的Autofac模組注入):
builder.Register( c => { var optionsBuilder = new DbContextOptionsBuilder<TestEFContext>(); //注入攔截器 optionsBuilder.AddInterceptors(new DbMasterSlaveCommandInterceptor(WriteConnect, ReadConnect)); //MaxBatchSize 處理批量操作BUG optionsBuilder.UseMysql(WriteConnect, b=>b.MaxBatchSize(1)); return optionsBuilder.Options; } ).As<DbContextOptions<TestEFContex>>().SingleInstance();
這樣就實現了通過攔截器實現讀寫分離.
2.通過攔截器實現SQL日誌記錄
同理,我們可以通過攔截器實現EF Core SQL語句的記錄與除錯
首先我們建立一個新的攔截器DBlogCommandInterceptor 如下:
public class DBlogCommandInterceptor : DbCommandInterceptor { //建立一個佇列記錄SQL執行時間 static readonly ConcurrentDictionary<DbCommand, DateTime> MStartTime = new ConcurrentDictionary<DbCommand, DateTime>(); private ILogger<DBlogCommandInterceptor> _logger { get; set; } //通過建構函式注入日誌 public DBlogCommandInterceptor(ILogger<DBlogCommandInterceptor> Logger) { _logger = Logger; } }
建立2個私有的方法,一個記錄執行開始時間,一個記錄SQL
//記錄SQL開始執行的時間 private void OnStart(DbCommand command) { MStartTime.TryAdd(command, DateTime.Now); } //通過_logger輸出日誌 private void Log(DbCommand command) { DateTime startTime; TimeSpan duration; //得到此command的開始時間 MStartTime.TryRemove(command, out startTime); if (startTime != default(DateTime)) { duration = DateTime.Now - startTime; } else { duration = TimeSpan.Zero; } var parameters = new StringBuilder(); //迴圈獲取執行語句的引數值 foreach (DbParameter param in command.Parameters) { parameters.AppendLine(param.ParameterName + " " + param.DbType + " = " + param.Value); } _logger.LogInformation("{starttime}開始執行SQL語句:{sql},引數:{canshu},執行時間{readtime}", startTime.ToString(), command.CommandText, parameters.ToString(), duration.TotalSeconds); }
最後過載攔截器的方法:
public override InterceptionResult<int> NonQueryExecuting(DbCommand command, CommandEventData eventData, InterceptionResult<int> result) { OnStart(command); return base.NonQueryExecuting(command, eventData, result); } public override Task<InterceptionResult<int>> NonQueryExecutingAsync(DbCommand command, CommandEventData eventData, InterceptionResult<int> result, CancellationToken cancellationToken = default) { OnStart(command); return base.NonQueryExecutingAsync(command, eventData, result, cancellationToken); } public override int NonQueryExecuted(DbCommand command, CommandExecutedEventData eventData, int result) { Log(command); return base.NonQueryExecuted(command, eventData, result); } public override Task<int> NonQueryExecutedAsync(DbCommand command, CommandExecutedEventData eventData, int result, CancellationToken cancellationToken = default) { Log(command); return base.NonQueryExecutedAsync(command, eventData, result, cancellationToken); } public override InterceptionResult<object> ScalarExecuting(DbCommand command, CommandEventData eventData, InterceptionResult<object> result) { OnStart(command); return base.ScalarExecuting(command, eventData, result); } public override Task<InterceptionResult<object>> ScalarExecutingAsync(DbCommand command, CommandEventData eventData, InterceptionResult<object> result, CancellationToken cancellationToken = default) { OnStart(command); return base.ScalarExecutingAsync(command, eventData, result, cancellationToken); } public override object ScalarExecuted(DbCommand command, CommandExecutedEventData eventData, object result) { Log(command); return base.ScalarExecuted(command, eventData, result); } public override Task<object> ScalarExecutedAsync(DbCommand command, CommandExecutedEventData eventData, object result, CancellationToken cancellationToken = default) { Log(command); return base.ScalarExecutedAsync(command, eventData, result, cancellationToken); } public override InterceptionResult<DbDataReader> ReaderExecuting(DbCommand command, CommandEventData eventData, InterceptionResult<DbDataReader> result) { OnStart(command); return base.ReaderExecuting(command, eventData, result); } public override Task<InterceptionResult<DbDataReader>> ReaderExecutingAsync(DbCommand command, CommandEventData eventData, InterceptionResult<DbDataReader> result, CancellationToken cancellationToken = default) { OnStart(command); return base.ReaderExecutingAsync(command, eventData, result, cancellationToken); } public override Task<DbDataReader> ReaderExecutedAsync(DbCommand command, CommandExecutedEventData eventData, DbDataReader result, CancellationToken cancellationToken = default) { Log(command); return base.ReaderExecutedAsync(command, eventData, result, cancellationToken); } public override DbDataReader ReaderExecuted(DbCommand command, CommandExecutedEventData eventData, DbDataReader result) { Log(command); return base.ReaderExecuted(command, eventData, result); }
這樣,我們就實現了通過攔截器實現SQL日誌記錄~效果如下:
除錯SQL語句就方便了很多~