204 Commits (8756d37bfca69aa072fcd959ad3dff2586ecdb2e)

Author	SHA1	Message	Date
Nick Vyzas	ab87981484	Adding optional lock option to destroy_MyConn_from_pool	8 years ago
René Cannaò	703ee8ef0f	Reset connections instead of dropping them when mysql-max_stmts_per_connection is reached ... Also fixed a memory leak associate with this feature. Cleaned up a bit of code.	8 years ago
René Cannaò	69654afa17	Do not dump replication hostgroups if hostgroup_manager_verbose=0 #1204	8 years ago
René Cannaò	4fb9e3f88b	Dump in error log status of mysql_servers during reconfiguration ... Close issue #1255	8 years ago
René Cannaò	ded88b1f2d	Merge branch 'v1.4.4-connthr' into v1.4.4	8 years ago
René Cannaò	e75d474878	Added mutex in replication_lag_action()	9 years ago
René Cannaò	18f6826ef7	Added new variable to limit the number of backend connections per second ... Variable name is `mysql-throttle_connections_per_sec_to_hostgroup` . Curently it is a global variable and limits the number of new connections per hostgroup, and not per specific node. For example, if mysql-throttle_connections_per_sec_to_hostgroup=100, no more than 100 new connections can be created on any hostgroup no matter the number of servers in that hostgroup. The default is very high (1000000) thus not changing default behaviour. Tuning this variable allows to control and throttle connections spikes to the backend servers. Added also new status variable `Server_Connections_delayed`. This is a counter of how many times Hostgroup Manager didn't return a connection because the limit was reached. It is worth to note that a single client request could make multiple requests, therefore this variable counts the number of time a new connection wasn't created and not how many requests were delayed.	9 years ago
René Cannaò	ec4049ac1e	Optimization on connections reset ... Limit the size of the queue #1185 Reuse the same user #1186	9 years ago
René Cannaò	b688120b25	Memory corruption on STMT resetted twice ... Bug #1197. Caused by PS resetted during change user, and then again when resetting the connection	9 years ago
René Cannaò	c9d3d03f32	Cluster: better handling of mysql_servers.status	9 years ago
René Cannaò	ed802e9f5d	Cluster: typo prevents checksum on mysql_servers	9 years ago
René Cannaò	a22eb8b395	No connection from pool for fast_forward #1157 ... If a user is configured with fast_forward, no connection should be taken from the connection pool. Yet the Hostgroup Manager should know about such connection.	9 years ago
René Cannaò	e3c4cb9964	A lot of changes all related to Cluster ... In `MySQL_Cluster` class added functions to sync with remote node, like: - `pull_mysql_query_rules_from_peer()` - `pull_mysql_servers_from_peer()` - `pull_mysql_users_from_peer()` - `pull_proxysql_servers_from_peer()` Added 8 new global variables in Admin. 4 variables determine after how many different checks the remote configuration will be synced: - cluster_mysql_query_rules_diffs_before_sync - cluster_mysql_servers_diffs_before_sync - cluster_mysql_users_diffs_before_sync - cluster_proxysql_servers_diffs_before_sync 4 variables determine if after a remote sync the changes need to be written to disk: - cluster_mysql_query_rules_save_to_disk - cluster_mysql_servers_save_to_disk - cluster_mysql_users_save_to_disk - cluster_proxysql_servers_save_to_disk Table `proxysql_servers` is now automatically loaded from disk to memory and into runtime at startup. Added new Admin's command `LOAD PROXYSQL SERVERS FROM CONFIG` to load `proxysql_servers` from config file to memory (not runtime). Internal structures with credentials in MySQL_Authentication moved from `unsorted_map` to `map` : this to ensure the right order when generating the checksum. Config file supports both `address` and `hostname` for `mysql_servers` and `proxysql_servers` , #1091 `ProxySQL_Admin::load_proxysql_servers_to_runtime()` now has a lock or no lock option, to avoid deadlock For now, Cluster module is quite verbose.	9 years ago
René Cannaò	2a76f95e31	Further enhancement on Cluster ... Extended class ProxySQL_Checksum_Value() in ProxySQL_Cluster module to support further metrics Implemeted `SELECT GLOBAL_CHECKSUM()` and relative tracking of global checksums Added variable `admin-cluster_check_status_frequency` to check peer's global status at regular intervals	9 years ago
René Cannaò	ee8110a515	Enhancement on Cluster solution ... This commit introduces: 2 new tables: * `runtime_checksums_values` : stores checksums of configurations in runtime. For now for `mysql_query_rules`, `mysql_servers` and `mysql_users` * `stats_proxysql_servers_checksums` : when clustering is enabled, it collects all metrics from `runtime_checksums_values` from all its peers 3 new global variables that defines it checksum needs to be generated during `LOAD ... TO RUNTIME` * `admin-checksum_mysql_query_rules` * `admin-checksum_mysql_servers` * `admin-checksum_mysql_users` ProxySQL Cluster connections now have timeouts: * 1 second timeout for CONNECT and WRITE * 60 seconds timeout for READ (useful for long poll)	9 years ago
René Cannaò	1c1faf210f	Some clean up	9 years ago
René Cannaò	dc3664ac33	Adding command PULL VERSION MYSQL SERVERS ... Experimental feature	9 years ago
René Cannaò	4149a5eff1	HGCU_thread_run() using password as username #1079 ... HGCU_thread_run() was incorrectly using monitor's password as username	9 years ago
René Cannaò	7ad078aebf	Final fix for #1061	9 years ago
René Cannaò	b78b687f46	Bug fix for #1061 ... Adding error handling in HGCU_thread_run() for `change_user_start()` and `change_user_cont()`	9 years ago
René Cannaò	27e28a53ad	Servers did not recover from RO=1 to RO=0 #1039 ... Status of the server was not changed if: * was OFFLINE_HARD * was incorrectly present in myhgm.mysql_servers Added also a mutex to serialize the read_only actions	9 years ago
René Cannaò	fc5a56eec3	Fix bug #1054, MySQL error 1203 ... Do not shun for error 1203: User %s already has more than 'max_user_connections' active connections	9 years ago
René Cannaò	080bb3d471	Big change to imporve how prepared statements work	9 years ago
René Cannaò	7823072dad	Unbalanced traffic with high weights #975 ... If total mysql_servers.weight per hostgroup was higher than 64K, traffic was unbalanced Conflicts: include/gen_utils.h lib/MySQL_HostGroups_Manager.cpp	9 years ago
René Cannaò	817cda3f12	New table stats_mysql_connection_pool_reset #753	9 years ago
René Cannaò	384c91b7ed	Merge branch 'v1.4.0-GR' into v1.4.0 ... Conflicts: src/main.cpp	9 years ago
René Cannaò	50e010bc54	Fix edge case during promotion #959 ... If a slave was shunned because replication was broken, when it becomes a master it was staying shunned	9 years ago
René Cannaò	2403b177ef	More support for Group Replication ... Enforced max_writers Enforced writer_is_also_reader Increased the number of entries in mysql_server_group_replication_log to 100 per node Lowered the minimum value for mysql-monitor_groupreplication_healthcheck_interval to 50ms Lowered the minimum value for mysql-monitor_groupreplication_healthcheck_timeout to 50ms	9 years ago
René Cannaò	752e3fd61e	Further improvement on forwarding #927 ... Introduced a timeout in HGCU_thread_run() Fixed a logical bug in MySQL_Protocol::process_pkt_COM_CHANGE_USER() that was causing to not authenticate user The absence of the timeout and the logicl bug was causing a lot of connections to be dropped and not really reused. But as now the connections are reused, they may present further bugs	9 years ago
René Cannaò	1259fb84e0	HGCU_thread_run corrupts connections #927	9 years ago
René Cannaò	c7cffc9ea7	More support for group replication	9 years ago
René Cannaò	63e86039e6	Added writer_hostgroup for GR	9 years ago
René Cannaò	669e6d3a20	Minor fixes for GR	9 years ago
René Cannaò	5817905df2	Merge branch 'v1.4.0' into v1.4.0-GR ... Conflicts: lib/MySQL_HostGroups_Manager.cpp lib/MySQL_Monitor.cpp lib/ProxySQL_Admin.cpp	9 years ago
René Cannaò	a6c53d0d12	Revert "Refactor header includes."	9 years ago
René Cannaò	8768acf5a5	Third commit for GR support	9 years ago
Maxim Bublis	30dbcc21a8	Refactor header includes. ... * reduce build time by 25%; * remove circular dependency on `cpp.h` and other headers. * fixes issue with `my_global.h` from mariadbclient which redefines `__attribute__`, that leads to broken STL and GCC intrinsics if `<algorithm>` wasn't included prior including this header; * fixes another issue with redefined `__attribute__` which leads to `_conn_exchange_t` not being aligned to cache line size; * removes `pthread_setstacksize` calls as `my_global.h` redefines `pthread_setstacksize` to be `pthread_dummy(0)` which led to stack size being never adjusted; * add missing include guards to some headers; * remove unused proxysql_hash.h header.	9 years ago
Maxim Bublis	043a321e55	Cleanup legacy codebase.	9 years ago
René Cannaò	7ea78394f0	Second commit for GR support	9 years ago
René Cannaò	3c2ced2a7d	Drafting GR support	9 years ago
René Cannaò	53d24f4e53	Speed up LOAD/SAVE mysql_servers #829 ... A further optimization in `MySQL_HostGroups_Manager::generate_mysql_servers_table()` makes `LOAD MYSQL SERVERS TO RUNTIME` around 35 times faster than original implementation. `SAVE MYSQL SERVERS TO MEMORY` is now over 10 times faster than the original implementation.	9 years ago
René Cannaò	ba7e37091b	Speed up LOAD MYSQL SERVERS TO RUNTIME #829 ... This commit is able to speed up LOAD MYSQL SERVERS TO RUNTIME up to 10x for very large mysql_servers table with tens of thousands of servers. Main improvements: * removed 1 unnecessary call to `generate_mysql_servers_table()` * replaced 1 call to `generate_mysql_servers_table()` with UPDATE statements when necessary * implemented new function `servers_add()` that replaces `server_add()` to perform bulk operation * replaced calls to `sqlite3_exec()` with proper parameters bindings * use of multirows INSERT statements (32 rows at the time) Miror improvements: * added new variable mysql-hostgroup_manager_verbose , 1 by default for backward compatibility and debugging * in not debug mode, access to HGM db has no shared cache * in not debug mode, tables in HGM have checks disabled Further note: * Once issue #797 is completed, `mysql_servers` in MyHGM can be simplified further	9 years ago
René Cannaò	86718882e0	Added 6 new status variables , #818 and #821 ... Added status variables: * Com_backend_change_user * Com_backend_init_db * Com_backend_set_names * Com_frontend_init_db * Com_frontend_set_names * Com_frontend_use_db	9 years ago
René Cannaò	de4f71bffa	Merge branch 'v1.3.0' into v1.3.1-dev ... Conflicts: deps/Makefile include/thread.h lib/MySQL_Monitor.cpp lib/thread.cpp src/main.cpp	9 years ago
René Cannaò	7680b8f752	Avoid double delete in read_only check	9 years ago
René Cannaò	e5531843ef	Fixed 2 memory leaks ... This should be the root cause of #766 and #796 This patch will be applied to v1.2.x , v1.3.0 and v1.3.1	9 years ago
René Cannaò	221e47315f	mysql_servers.comment not loaded at runtime #787 ... See also #782	9 years ago
René Cannaò	d0d764533d	Specify when a mysql_users.max_connections is reached #794 ... It also fixes: * bug in deps/Makefile related to cleanall * crashing bug in HGCU_thread_run()	9 years ago
René Cannaò	226f71d4c3	Connections are sent to purge thread if queue<100	9 years ago
René Cannaò	159176627d	Thread to reset connections	9 years ago
René Cannaò	799c7d411a	Merge pull request #737 from ton31337/feature/do_not_shunn_on_few_error_codes ... Do not shunt the server if max_user_connections is reached for user	10 years ago
Donatas Abraitis	0e062058e6	Do not shunt the server if max_user_connections is reached for user	10 years ago
René Cannaò	01395f93f2	Merge branch 'v1.2.3-644' into 1.3.0-alpha	10 years ago
René Cannaò	5330000348	Added variables to control prepared statements ... - mysql-max_stmts_per_connection - mysql-max_stmts_cache	10 years ago
René Cannaò	c987da5b03	Various changes ... - "Prepare" wasn't listed in SHOW PROCESSLIST - during ~MySQL_Connection(), all	10 years ago
René Cannaò	8ce083fb20	Testing pthread mutex for MHM #644 ... If MHM_PTHREAD_MUTEX is defined MySQL_HostGroups_Manager.h , MySQL_HostGroups_Manager() is compiled using pthread mutex instead od spin loop. As a lot of computation is protected by mutex, maybe it is better not to use spin loop.	10 years ago
René Cannaò	861b12fe52	Minor performance improvements ... PtrArray is now header only Moved some function of MySrvConnList() into header	10 years ago
René Cannaò	be063707e9	MySrvConnList::conns is now private ... This commit is a preparation to convert conns from an array to a map	10 years ago
René Cannaò	7d3814e5b9	Scalability improvements #644 ... Some minor yet important tweak to spinloop Replaced rand() with fastrand()	10 years ago
René Cannaò	f27734390e	Merge branch 'v1.2.2-RO-loop' into v1.2.2	10 years ago
René Cannaò	d214e2f60b	Fix bug #667	10 years ago
René Cannaò	12e1223125	Local connection cache #644 ... This is an attempt to improve scalability adding a local connection cache per thread	10 years ago
René Cannaò	79d82929a5	Changed severity of messages when node is shunned ... When a server is shunned because of errors, it is now logged as error When a server is shunned because of replication lag, it is now logged as warning	10 years ago
René Cannaò	e6621ef5b3	Fix bug in read_only_action due to comment #643	10 years ago
René Cannaò	54c13e1dc1	Copy comment in read_only_action ... Issue #643	10 years ago
René Cannaò	a9395cd087	Adding comment field per issue #643 ... Added: * mysql_query_rules.comment VARCHAR * mysql_replication_hostgroups.comment VARCHAR * mysql_servers.comment VARCHAR NOT NULL DEFAULT '' All these tables support online upgrade from previous version (without such fields) at startup. Introduced also a parser for escaping strings with single quotes: double quotes are allowed without extra escaping	10 years ago
René Cannaò	95948d7eb0	Fix for bug #641 ... A server is copied from reader hostgroup to writer hostgroup only if absent or OFFLINE_HARD in writer hostgroup	10 years ago
René Cannaò	e94df8264a	Revert last commit, and added comment	10 years ago
René Cannaò	dc804eccea	Fix bug in read_only_action()	10 years ago
René Cannaò	b8d3bd884c	Added more memory metrics #634 ... * mysql_backend_buffers_bytes * mysql_frontend_buffers_bytes * mysql_session_internal_bytes * ConnPool_memory_bytes	10 years ago
René Cannaò	a0c994ea43	Server was incorrectly set to SHUNNED ... Server was incorrectly set to SHUNNED without checking previous status See #627	10 years ago
René Cannaò	8c342ddf18	Incorrect status in shun_and_killall() ... MySQL_HostGroups_Manager::shun_and_killall() incorrectly set status=SHUNNED for server in OFFLINE_SOFT	10 years ago
René Cannaò	1126001dd8	Return SHUNNED if status=4 ... If status=4 , MySQL_HostGroups_Manager::dump_table_mysql_servers() should return SHUNNED Previous nothing was returned for status=4 (MYSQL_SERVER_STATUS_SHUNNED_REPLICATION_LAG) , caausing the crashing bug #619 dump_table_mysql_servers()	10 years ago
René Cannaò	77d2a9c1ad	Removed a compiler warning	10 years ago
René Cannaò	c11303b117	Servers were incorrectly brought back online ... Fix issue #566 Regression introducted in #531	10 years ago
René Cannaò	b19b21bb56	Fixed minor typo	10 years ago
René Cannaò	510dde94a6	Merge branch 'max_latency_ms' into v1.2.0	10 years ago
René Cannaò	0f70d24b13	Merge branch 'ssl' into v1.2.0	10 years ago
René Cannaò	fe12d26fe4	Implementation of mysql-connection_max_age_ms ... If a connection has a lifetime greated than mysql-connection_max_age_ms , it is dropped See #563	10 years ago
René Cannaò	f27bb873f0	MySQL_Monitor call to set_server_current_latency_us ... Fixed a wrong implementation in previous commit This commit also updates stats_mysql_connection_pool	10 years ago
René Cannaò	517c35f25f	Renamed function ... Renamed MySQL_HostGroups_Manager::set_server_current_latency_ms() to MySQL_HostGroups_Manager::set_server_current_latency_us()	10 years ago
René Cannaò	81e8239917	Connection pool considers current_latency_us ... When evaluating if a host can be used, current_latency_us is taken into consideration	10 years ago
René Cannaò	0fef6b544c	Added global variables mysql-default_max_latency_ms	10 years ago
René Cannaò	41ee4e5b85	Fixed typo in previous commit related to SSL and latency ... Also testing SSL connection with hardcoded certificates	10 years ago
René Cannaò	9019c6b3d5	Added columns use_ssl and max_latency_ms to mysql_servers ... These two new columns will be used to: - enable SSL on a per server basis - define maximum latency for a server to be consider usable: this will be a key element for multi DC support	10 years ago
René Cannaò	820dd6357e	Fixed major bug introduced in #545 ... The math was balancing connections was incorrect	10 years ago
René Cannaò	00caba00fc	Minor bug fixes on #544 and #545	10 years ago
René Cannaò	db33ee961e	Balance connections usage based on weight and used connections #545 ... First implementation	10 years ago
René Cannaò	fc9ab855da	First draft for bugs #543 ... When a server isn't responding to ping, it is flagged as shunned and all the connections need to be dropped. All the connections must be dropped before the server is brought back online	10 years ago
René Cannaò	751ba624a3	Auto tuning shun_recovery_time_sec ... See issues #530 and #531 Closes #530 Closes #531	10 years ago
René Cannaò	66280b1bd3	Fixed a race condition in LOAD/SAVE MYSQL SERVERS	10 years ago
René Cannaò	c6b7f6db52	Merge branch 'servers_version' into beta7	10 years ago
René Cannaò	a3dafd830e	Added status variable Servers_table_version ... This status variable is incremented every time mysql_servers table is reloaded	10 years ago
René Cannaò	eb012cb782	Removed a not relevant verbose line	10 years ago
René Cannaò	376b037585	Added support for runtime mysql servers tables ... Issue #488 - runtime_mysql_replication_hostgroups - runtime_mysql_servers	10 years ago
René Cannaò	dca5dd7e8f	Do not shun for error in replication lag ... error in retrieving replication lag does not shun a slave. See #479	10 years ago
René Cannaò	2ea4149b0b	Fixed accounting of Server_Connections_connected	10 years ago
René Cannaò	efbdf3b617	Try to free mysql buffers when possible ... Free mysql->net.buff when beyond 64KB	10 years ago
René Cannaò	89c7fad16e	Fixed bug in accounting Server_Connections ... Server_Connections_connected and Server_Connections_created were inverted	10 years ago
René Cannaò	abef155790	Prevent "Unknown database" to shun backends ... Issue #464 Attempt to connect to a not existing schema caused the backend to be shunned. This patch fix this issue. Closes #464	10 years ago