在上一篇文章《使用ProxySQL實現MySQL Group Replication的故障轉移、讀寫分離（一） 》 中，已經完成了MGR+ProxySQL集群的搭建，也測試了ProxySQL實現業務層面的故障轉移，接下來繼續測試讀寫分離。
王國維大師筆下的人生三境界：
第一重境界：昨夜西風凋碧樹。獨上高樓，望盡天涯路；
第二重境界：衣帶漸寬終不悔，為伊消得人憔悴；
第三重境界：眾里尋他千百度，驀然回首，那人卻在燈火闌珊處。
作為一個一根筋的學渣程序員,我還沒能想透徹。但是數據庫讀寫分離的三境界卻有了一定的了解，我們不妨來看一看MySQL數據庫讀寫分離的三境界。
第一重境界：人工實現讀寫分離。通過IP、端口讀寫分離，業務層面人工識別讀寫語句，然后將其分配到不同的主機，實現讀寫分離；
第二重境界：正則實現讀寫分離。通過路由中間件識別SQL語句，通過正則表達式匹配SQL語句，然后根據匹配結果分發到不同的主機；
第三重境界：識別TOP SQL，將高負載SQL分發到不同的主機；
（一）第一境界：人工實現讀寫分離
通過IP、端口讀寫分離，業務層面人工識別讀寫語句，然后使用不同的連接數據庫配置信息，將其分配到不同的主機，實現讀寫分離。在ProxySQL里面，我們是通過端口來實現讀寫分離的。具體操作如下：
STEP1：配置ProxySQL在兩個端口上偵聽，並且重新啟動ProxySQL
mysql -uadmin -padmin -h127.0.0.1 -P6032 mysql> SET mysql-interfaces='0.0.0.0:6401;0.0.0.0:6402'; -- save it on disk and restart proxysql mysql> SAVE MYSQL VARIABLES TO DISK; mysql> PROXYSQL RESTART;
STEP2：配置路由規則，通過端口將請求分發到不同的組
mysql> INSERT INTO mysql_query_rules (rule_id,active,proxy_port,destination_hostgroup,apply) VALUES (1,1,6401,10,1), (3,1,6402,20,1); mysql> LOAD MYSQL QUERY RULES TO RUNTIME; mysql> SAVE MYSQL QUERY RULES TO DISK;
這樣，通過6401端口訪問數據庫的請求就會被轉發到組1(寫組)中，通過6402端口訪問數據庫的請求會被轉發到組3(讀組)中，從而實現讀寫分離，具體使用6401端口還是6402端口訪問數據庫，取決於開發人員人工識別SQL的讀寫特性。
（二）第二境界：使用正則表達式實現讀寫分離
通過路由中間件識別SQL語句，通過正則表達式匹配SQL語句，然后根據匹配結果分發到不同的主機。操作過程如下
STEP1:為避免干擾測試，刪除之前定義的規則
DELETE FROM mysql_query_rules;
STEP2:定義新的讀寫分離規則
INSERT INTO mysql_query_rules (rule_id,active,match_digest,destination_hostgroup,apply) VALUES(1,1,'^SELECT.*FOR UPDATE$',1,1); INSERT INTO mysql_query_rules (rule_id,active,match_digest,destination_hostgroup,apply) VALUES(2,1,'^SELECT',3,1); LOAD MYSQL QUERY RULES TO RUNTIME; SAVE MYSQL QUERY RULES TO DISK;
現在，ProxySQL的路由規則為：
- SELECT FOR UPDATE操作將被路由到組1（寫組）;
- 其它的SELECT語句將被路由到組3(讀組);
- 其它的路由到默認組，即組1。
這里對使用正則表達式方式進行測試，整個過程如下：
（1）測試之前讀寫組信息修改
-- 根據組的規則：最多1個寫節點，其余的寫節點放入備用寫組。目前我們可以看到節點192.168.10.13是寫節點，其余2個節點是備用寫節點，沒有讀節點 mysql> select * from mysql_group_replication_hostgroups; +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | writer_hostgroup | backup_writer_hostgroup | reader_hostgroup | offline_hostgroup | active | max_writers | writer_is_also_reader | max_transactions_behind | comment | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | 1 | 2 | 3 | 4 | 1 | 1 | 0 | 100 | NULL | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ 1 row in set (0.00 sec) mysql> mysql> select * from runtime_mysql_servers; +--------------+---------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ | hostgroup_id | hostname | port | gtid_port | status | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment | +--------------+---------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ | 1 | 192.168.10.13 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | | 2 | 192.168.10.12 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | | 2 | 192.168.10.11 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | +--------------+---------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ 3 rows in set (0.01 sec) -- 為了實現讀寫分離，需要有讀節點，我們可以修改writer_is_also_reader參數，讓backup_writer_hostgroup中的節點既做備用寫節點，又做讀節點 mysql> update mysql_group_replication_hostgroups set writer_is_also_reader = 2 ; Query OK, 1 row affected (0.00 sec) mysql> select * from mysql_group_replication_hostgroups; +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | writer_hostgroup | backup_writer_hostgroup | reader_hostgroup | offline_hostgroup | active | max_writers | writer_is_also_reader | max_transactions_behind | comment | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | 1 | 2 | 3 | 4 | 1 | 1 | 2 | 100 | NULL | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ 1 row in set (0.00 sec) mysql> mysql> select * from runtime_mysql_group_replication_hostgroups; +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | writer_hostgroup | backup_writer_hostgroup | reader_hostgroup | offline_hostgroup | active | max_writers | writer_is_also_reader | max_transactions_behind | comment | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | 1 | 2 | 3 | 4 | 1 | 1 | 0 | 100 | NULL | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ 1 row in set (0.00 sec) --需要生效、永久保存mysql server配置 mysql> load mysql servers to runtime; Query OK, 0 rows affected (0.01 sec) mysql> save mysql servers to disk; Query OK, 0 rows affected (0.03 sec) mysql> select * from runtime_mysql_group_replication_hostgroups; +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | writer_hostgroup | backup_writer_hostgroup | reader_hostgroup | offline_hostgroup | active | max_writers | writer_is_also_reader | max_transactions_behind | comment | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ | 1 | 2 | 3 | 4 | 1 | 1 | 2 | 100 | NULL | +------------------+-------------------------+------------------+-------------------+--------+-------------+-----------------------+-------------------------+---------+ 1 row in set (0.01 sec) -- 最終mysql server的組信息如下 mysql> select * from runtime_mysql_servers; +--------------+---------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ | hostgroup_id | hostname | port | gtid_port | status | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment | +--------------+---------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ | 1 | 192.168.10.13 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | | 3 | 192.168.10.12 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | | 3 | 192.168.10.11 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | | 2 | 192.168.10.11 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | | 2 | 192.168.10.12 | 3306 | 0 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | +--------------+---------------+------+-----------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ 5 rows in set (0.00 sec)
（2）導入規則
-- 為避免測試干擾，先刪除之前的規則 DELETE FROM mysql_query_rules; -- 導入規則 INSERT INTO mysql_query_rules (rule_id,active,match_digest,destination_hostgroup,apply) VALUES(1,1,'^SELECT.*FOR UPDATE$',1,1); INSERT INTO mysql_query_rules (rule_id,active,match_digest,destination_hostgroup,apply) VALUES(2,1,'^SELECT',3,1); -- 生效、保存規則 LOAD MYSQL QUERY RULES TO RUNTIME; SAVE MYSQL QUERY RULES TO DISK;
（3）測試規則是否生效
測試SQL語句：
mysql -uusera -p123456 -h192.168.10.10 -P6033 -- 寫測試 insert into testdb.test01 values(3,'c'); -- 讀測試 SELECT * from testdb.test01; -- 正則大小寫測試 select * from testdb.test01; -- select for update測試 SELECT * from testdb.test01 FOR UPDATE; select * from testdb.test01 FOR UPDATE; exit;
ProxySQL將SQL語句分發到哪一台主機上執行，可以查看統計視圖：stats_mysql_query_digest和stats_mysql_query_digest_reset。兩個表的內容和結構相同，但是查詢stats_mysql_query_digest_reset表會自動將內部統計信息重置為零，即執行了stats_mysql_query_digest_reset的查詢后，2個表的數據都會被完全清除。這里我們直接使用stats_mysql_query_digest_reset來查詢上面的測試：
mysql> select hostgroup,schemaname,username,digest_text,count_star from stats_mysql_query_digest_reset; +-----------+--------------------+----------+----------------------------------------+------------+ | hostgroup | schemaname | username | digest_text | count_star | +-----------+--------------------+----------+----------------------------------------+------------+ | 1 | information_schema | usera | SELECT * from testdb.test01 FOR UPDATE | 1 | | 3 | information_schema | usera | select * from testdb.test01 | 1 | | 3 | information_schema | usera | SELECT * from testdb.test01 | 1 | | 1 | information_schema | usera | select * from testdb.test01 FOR UPDATE | 1 | | 1 | information_schema | usera | insert into testdb.test01 values(?,?) | 1 | | 1 | information_schema | usera | select @@version_comment limit ? | 1 | +-----------+--------------------+----------+----------------------------------------+------------+ 6 rows in set (0.00 sec)
可以看到，正則表達式規則不區分大小寫，並且根據匹配規則，已經將SQL發到了對應的主機上執行。
個人覺得基於正則表達式路由SQL語句到不同主機執行已經十分智能了，然而ProxySQL官方並不建議這么干，因為我們無法准確知道各類型的SQL語句的開銷，從而可能會導致流量分布不均。
接下來我們來看看ProxySQL推薦的方法，基於正則表達式和摘要進行讀寫拆分。
（三）第三境界：使用正則表達式和digest實現讀寫分離
以下是ProxySQL推薦的有效設置讀寫分離的配置過程：
(1)配置ProxySQL以將所有流量僅發送到一個MySQL主節點，寫和讀都發送到一個節點；
(2)檢查stats_mysql_query_digest哪些是最昂貴的SELECT語句；
(3)確定哪些昂貴的語句應移至讀節點；
(4)配置mysql_query_rules（創建規則）以僅將昂貴的SELECT語句發送給讀者
總之，想法非常簡單：僅發送那些你想發送的SQL給讀節點，而不是發送所有SELECT語句。
我們來整理一下整個過程：
STEP1:去除規則，讓所有SQL語句都在默認組上執行
mysql> delete from mysql_query_rules; Query OK, 2 rows affected (0.00 sec) mysql> LOAD MYSQL QUERY RULES TO RUNTIME; Query OK, 0 rows affected (0.00 sec) mysql> SAVE MYSQL QUERY RULES TO DISK; Query OK, 0 rows affected (0.01 sec)
STEP2:查找最昂貴的SQL
假設目前所有讀寫操作都在同一台機器上執行，且執行了很久，讀寫比例都具有代表性，我們可以使用stats_mysql_query_digest查找最昂貴的SQL，可以多維度進行查找。
（1）查找查詢總耗時最多的5個SQL
mysql> SELECT digest,SUBSTR(digest_text,0,25),count_star,sum_time FROM stats_mysql_query_digest WHERE digest_text LIKE 'SELECT%' ORDER BY sum_time DESC LIMIT 5; +--------------------+--------------------------+------------+----------+ | digest | SUBSTR(digest_text,0,25) | count_star | sum_time | +--------------------+--------------------------+------------+----------+ | 0xBF001A0C13781C1D | SELECT c FROM sbtest1 WH | 9594 | 9837782 | | 0xC4771449056AB3AC | SELECT c FROM sbtest14 W | 9984 | 9756595 | | 0xD84E4E04982951C1 | SELECT c FROM sbtest9 WH | 9504 | 9596185 | | 0x9B090963F41AD781 | SELECT c FROM sbtest10 W | 9664 | 9530433 | | 0x9AF59B998A3688ED | SELECT c FROM sbtest2 WH | 9744 | 9513180 | +--------------------+--------------------------+------------+----------+ 5 rows in set (0.00 sec)
（2）查看執行次數最多的5個SQL語句
mysql> SELECT digest,SUBSTR(digest_text,0,25),count_star,sum_time FROM stats_mysql_query_digest WHERE digest_text LIKE 'SELECT%' ORDER BY count_star DESC LIMIT 5; +--------------------+--------------------------+------------+----------+ | digest | SUBSTR(digest_text,0,25) | count_star | sum_time | +--------------------+--------------------------+------------+----------+ | 0xC4771449056AB3AC | SELECT c FROM sbtest14 W | 9984 | 9756595 | | 0x9AF59B998A3688ED | SELECT c FROM sbtest2 WH | 9744 | 9513180 | | 0x9B090963F41AD781 | SELECT c FROM sbtest10 W | 9664 | 9530433 | | 0x03744DC190BC72C7 | SELECT c FROM sbtest5 WH | 9604 | 9343514 | | 0x1E7B7AC5611F30C2 | SELECT c FROM sbtest6 WH | 9594 | 9245838 | +--------------------+--------------------------+------------+----------+
（3）查看平均執行時間最長的5個SQL語句
mysql> SELECT digest,SUBSTR(digest_text,0,25),count_star,sum_time, sum_time/count_star as avg_time FROM stats_mysql_query_digest WHERE digest_text LIKE 'SELECT%' ORDER BY avg_time DESC LIMIT 5; +--------------------+--------------------------+------------+----------+----------+ | digest | SUBSTR(digest_text,0,25) | count_star | sum_time | avg_time | +--------------------+--------------------------+------------+----------+----------+ | 0x0DCAF47B4A363A7A | SELECT * from testdb.tes | 1 | 11400 | 11400 | | 0x2050E81DB9C7038E | select * from testdb.tes | 1 | 10817 | 10817 | | 0xF340A73F6EDA5B20 | SELECT c FROM sbtest11 W | 964 | 1726994 | 1791 | | 0xC867A28C90150A81 | SELECT DISTINCT c FROM s | 929 | 1282699 | 1380 | | 0x283AA9863F85EFC8 | SELECT DISTINCT c FROM s | 963 | 1318362 | 1369 | +--------------------+--------------------------+------------+----------+----------+
（4）查看平均執行時間最長的5個SQL語句，且滿足平均執行時間大於1s，並顯示該SQL執行時間占所有SQL執行時間的百分比
SELECT digest,SUBSTR(digest_text,0,25),count_star,sum_time,sum_time/count_star as avg_time,round(sum_time/1000000*100/(SELECT sum(sum_time/1000000) FROM stats_mysql_query_digest ),3) as pct FROM stats_mysql_query_digest WHERE digest_text LIKE 'SELECT%' AND sum_time/count_star > 1000000 ORDER BY avg_time DESC LIMIT 5;
說明：在測試該語句時，是使用sysbench壓測出來的數據，發現存在一個sum_time非常大的SQL，導致在求sum(sum_time)時返回NULL值，故先做了預處理，把sum_time/1000000變為進行計算。
STEP3:結合digest和正則表達式實現路由
我們先觀察一下，未使用路由規則時候的流量分布，可以看到，所有流量都到了hostgroup1
mysql> select hostgroup,schemaname,username,digest_text,count_star from stats_mysql_query_digest_reset; +-----------+--------------------+----------+---------------------------------------------------------------------+------------+ | hostgroup | schemaname | username | digest_text | count_star | +-----------+--------------------+----------+---------------------------------------------------------------------+------------+ | 1 | information_schema | usera | SET PROFILING = ? | 1 | | 1 | information_schema | usera | SHOW DATABASES | 3 | | 1 | information_schema | usera | SHOW VARIABLES LIKE ?; | 2 | | 1 | information_schema | usera | SET NAMES utf8mb4 | 3 | | 1 | tssysbench | usera | INSERT INTO sbtest15 (id, k, c, pad) VALUES (?, ?, ?, ?) | 1285 | | 1 | tssysbench | usera | INSERT INTO sbtest14 (id, k, c, pad) VALUES (?, ?, ?, ?) | 1309 | | 1 | tssysbench | usera | INSERT INTO sbtest13 (id, k, c, pad) VALUES (?, ?, ?, ?) | 1303 | | 1 | tssysbench | usera | INSERT INTO sbtest12 (id, k, c, pad) VALUES (?, ?, ?, ?) | 1240 | | 1 | tssysbench | usera | UPDATE sbtest3 SET k=k+? WHERE id=? | 1280 | | 1 | tssysbench | usera | UPDATE sbtest2 SET k=k+? WHERE id=? | 1280 | | 1 | tssysbench | usera | UPDATE sbtest1 SET k=k+? WHERE id=? | 1219 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest15 WHERE id BETWEEN ? AND ? ORDER BY c | 1207 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest14 WHERE id BETWEEN ? AND ? ORDER BY c | 1262 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest11 WHERE id BETWEEN ? AND ? ORDER BY c | 1227 |
插入路由規則：
-- 根據digest插入規則，匹配特定的SQL語句 INSERT INTO mysql_query_rules (rule_id,active,digest,destination_hostgroup,apply) VALUES(1,1,'0x0DCAF47B4A363A7A',3,1); INSERT INTO mysql_query_rules (rule_id,active,digest,destination_hostgroup,apply) VALUES(2,1,'0x63F9BD89D906209B',3,1); INSERT INTO mysql_query_rules (rule_id,active,digest,destination_hostgroup,apply) VALUES(3,1,'0x10D8D9CC551E199B',3,1); INSERT INTO mysql_query_rules (rule_id,active,digest,destination_hostgroup,apply) VALUES(4,1,'0xC867A28C90150A81',3,1); INSERT INTO mysql_query_rules (rule_id,active,digest,destination_hostgroup,apply) VALUES(5,1,'0x283AA9863F85EFC8',3,1); INSERT INTO mysql_query_rules (rule_id,active,digest,destination_hostgroup,apply) VALUES(6,1,'0x16BD798E66615299',3,1); -- 根據正則表達式插入規則，匹配所有SELECT 開頭的語句 INSERT INTO mysql_query_rules (rule_id,active,match_digest,destination_hostgroup,apply) VALUES(7,1,'^SELECT COUNT\(\*\)',3,1); -- 使規則生效、保存 LOAD MYSQL QUERY RULES TO RUNTIME; SAVE MYSQL QUERY RULES TO DISK;
STEP4:使用sysbench查詢，再次查看流量分布，可以看到，符合路由條件的SQL語句已經轉移到了hostgroup3執行。
mysql> select hostgroup,schemaname,username,digest_text,count_star from stats_mysql_query_digest_reset; +-----------+------------+----------+---------------------------------------------------------------------+------------+ | hostgroup | schemaname | username | digest_text | count_star | +-----------+------------+----------+---------------------------------------------------------------------+------------+ | 1 | tssysbench | usera | UPDATE sbtest3 SET k=k+? WHERE id=? | 863 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest14 WHERE id BETWEEN ? AND ? ORDER BY c | 841 | | 3 | tssysbench | usera | SELECT DISTINCT c FROM sbtest13 WHERE id BETWEEN ? AND ? ORDER BY c | 765 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest12 WHERE id BETWEEN ? AND ? ORDER BY c | 837 | | 3 | tssysbench | usera | SELECT DISTINCT c FROM sbtest11 WHERE id BETWEEN ? AND ? ORDER BY c | 813 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest10 WHERE id BETWEEN ? AND ? ORDER BY c | 861 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest9 WHERE id BETWEEN ? AND ? ORDER BY c | 835 | | 3 | tssysbench | usera | SELECT DISTINCT c FROM sbtest8 WHERE id BETWEEN ? AND ? ORDER BY c | 823 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest6 WHERE id BETWEEN ? AND ? ORDER BY c | 834 | | 1 | tssysbench | usera | UPDATE sbtest5 SET c=? WHERE id=? | 870 | | 3 | tssysbench | usera | SELECT DISTINCT c FROM sbtest4 WHERE id BETWEEN ? AND ? ORDER BY c | 802 | | 1 | tssysbench | usera | UPDATE sbtest1 SET c=? WHERE id=? | 835 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest3 WHERE id BETWEEN ? AND ? ORDER BY c | 838 | | 1 | tssysbench | usera | SELECT DISTINCT c FROM sbtest2 WHERE id BETWEEN ? AND ? ORDER BY c | 885
至此，以實現根據負載進行流量分發。
================================================================================================================
附1：讀寫分離路由規則表解析
讀寫分離路由解析信息存放在mysql_query_rules表中，表的語法如下：
CREATE TABLE mysql_query_rules ( rule_id INTEGER PRIMARY KEY AUTOINCREMENT NOT NULL, active INT CHECK (active IN (0,1)) NOT NULL DEFAULT 0, username VARCHAR, schemaname VARCHAR, flagIN INT CHECK (flagIN >= 0) NOT NULL DEFAULT 0, client_addr VARCHAR, proxy_addr VARCHAR, proxy_port INT, digest VARCHAR, match_digest VARCHAR, match_pattern VARCHAR, negate_match_pattern INT CHECK (negate_match_pattern IN (0,1)) NOT NULL DEFAULT 0, re_modifiers VARCHAR DEFAULT 'CASELESS', flagOUT INT CHECK (flagOUT >= 0), replace_pattern VARCHAR CHECK(CASE WHEN replace_pattern IS NULL THEN 1 WHEN replace_pattern IS NOT NULL AND match_pattern IS NOT NULL THEN 1 ELSE 0 END), destination_hostgroup INT DEFAULT NULL, cache_ttl INT CHECK(cache_ttl > 0), cache_empty_result INT CHECK (cache_empty_result IN (0,1)) DEFAULT NULL, reconnect INT CHECK (reconnect IN (0,1)) DEFAULT NULL, timeout INT UNSIGNED, retries INT CHECK (retries>=0 AND retries <=1000), delay INT UNSIGNED, next_query_flagIN INT UNSIGNED, mirror_flagOUT INT UNSIGNED, mirror_hostgroup INT UNSIGNED, error_msg VARCHAR, OK_msg VARCHAR, sticky_conn INT CHECK (sticky_conn IN (0,1)), multiplex INT CHECK (multiplex IN (0,1,2)), gtid_from_hostgroup INT UNSIGNED, log INT CHECK (log IN (0,1)), apply INT CHECK(apply IN (0,1)) NOT NULL DEFAULT 0, comment VARCHAR)
重要列的含義如下：
- rule_id ：規則的id，是主鍵，具有唯一非空特性，規則匹配時，按照rule_id從小到大匹配；
- active ：規則是否啟用，1代表啟用；
- username: : 匹配來自特定用戶的流量；
- client_addr ：匹配來自特定客戶端的流量；
- proxy_addr : 匹配特定本地IP上的傳入流量；
- proxy_port : 匹配特定本地端口上的傳入流量，具體見上面使用端口進行讀寫分離的方案；
- digest : 將查詢與特定摘要匹配，每個相同的SQL文本都會生成一個唯一的diagst碼(類似Oracle的sql_id)，按照碼進行匹配；
- match_digest ：將查詢摘要與正則表達式匹配；
- match_pattern：將查詢文本與正則表達式匹配；
- destination_hostgroup：將匹配的查詢路由到該主機組，除非存在已啟動的事務並且已登錄的用戶將transaction_persistent標志設置為1（請
參見表mysql_users），否則將發生這種情況。 - cache_ttl ：查詢結果緩存保留的時間(單位：s)；
- timeout ：執行匹配或重寫的查詢的最大超時（以毫秒為單位）。如果查詢的運行時間超過特定閾值，則會自動終止該查詢。如果未指定超時，則
mysql-default_query_timeout應用全局變量 - retries : 在執行查詢檢測到失敗的情況下，重新執行查詢的次數
- apply : 如果這只為1，則不再匹配后面的查詢規則。
附2：本次實驗用到的sysbench腳本
-- 准備階段
sysbench /usr/share/sysbench/oltp_read_write.lua --mysql-host=192.168.10.10 --mysql-port=6033 --mysql-user=usera --mysql-password='123456' --mysql-db=tssysbench --db-driver=mysql --tables=15 --table-size=50000 --report-interval=10 --threads=4 --time=120 prepare
-- 測試階段 sysbench /usr/share/sysbench/oltp_read_write.lua --mysql-host=192.168.10.10 --mysql-port=6033 --mysql-user=usera --mysql-password='123456' --mysql-db=tssysbench --db-driver=mysql --tables=15 --table-size=500000 --report-interval=10 --threads=4 --time=120 run
-- 清除階段 sysbench /usr/share/sysbench/oltp_read_write.lua --mysql-host=192.168.10.10 --mysql-port=6033 --mysql-user=usera --mysql-password='123456' --mysql-db=tssysbench --db-driver=mysql --tables=15 --table-size=500000 --report-interval=10 --threads=4 --time=120 cleanup
【完】