16 Commits (97eab1bbf0ce3c7c453e8a8bb96b2bd564e4758f)

Author	SHA1	Message	Date
Rahim Kanji	6909011b67	perf(partition): promote longest-waiting B session via max_connect_time fold ... Under sustained contention ProcessAllSessions_Partition leaves the B (CONNECTING_SERVER) band unordered, so the pass can serve newer waiters ahead of older ones until they hit connect_timeout_server_max and abort. Track the oldest waiter during the classifier's is_B arm (no second pass) and swap it to running_end. Key on max_connect_time, which the is_B test already loads -- the min-compare is a register compare with no extra memory access, and the smallest value is the waiter closest to the timeout abort. Gated by |B| > PARTITION_FAIRNESS_MIN_B (=4) and partition_active.	3 weeks ago
Rahim Kanji	53eea80aaf	feat(partition-gate): suppress low-volume ticks via min-attempts floor ... Add PARTITION_GATE_MIN_ATTEMPTS (=4). When a tick has fewer pool acquires than this, leave both gate state and streak counter untouched. Filters the "2/2 NULL = 100% stressed" noise case without drifting the gate OFF on transient idle ticks (which would happen if low- volume ticks were treated as not-stressed). Pairs with the existing mysql_sessions->len > 3 guard at the partition call site.	4 weeks ago
Rahim Kanji	878ba3b1e6	refactor(partition-gate): consolidate into Base_Thread; drain every iteration; widen ratio to uint64_t ... Three closely related changes: * Move the four partition_* members and note_pool_attempt() from MySQL_Thread / PgSQL_Thread into Base_Thread. Eliminates 22 lines of duplicated gate logic across the two protocol implementations. * Run update_partition_gate() once per process_all_sessions outer iteration, before the sess_sort / len > 3 / partition_active check. Per-tick counters can no longer accumulate stale across ticks where partition is skipped (sess_sort off, IDLE_THREADS maintenance thread, or session count below threshold), which in 5b5b937 could cause spurious gate flips after sess_sort toggled back on. * Widen the ratio comparison to uint64_t so the multiplications cannot overflow unsigned int (reachable in combination with the staleness issue above). Function-local HI_NUM/HI_DEN/STREAK locals are promoted to named static constexpr class members (PARTITION_GATE_NULL_RATIO_NUM/DEN/STREAK). No new behavior introduced -- purely correctness fixes plus a behavior-preserving refactor.	4 weeks ago
Rahim Kanji	ca202ceba1	Refactor: partition sessions by backend state instead of sorting ... Replace ProcessAllSessions_SortingSessions with ProcessAllSessions_Partition. The old function did an implicit one-pass swap-sort over mysql_sessions by max_connect_time; the new one explicitly partitions sessions into three contiguous blocks in pdata according to their backend state: [0, running_end) A: running a query against the backend [running_end, idle_begin) B: acquiring a backend (max_connect_time != 0) [idle_begin, len) C: idle (no backend, or holding a conn while parked in WAITING_CLIENT_DATA)	1 month ago
Rahim Kanji	9e4b76d6d1	* Refactored PgSQL_Session::handler___status_WAITING_CLIENT_DATA___STATE_SLEEP___PGSQL_QUERY_qpo. ... * Fixed a crash occurring during session destruction. * Query rules will now apply only to the first message in an extended query frame. * OK message will apply to execute message. * Query rewrite, error messages, and large packet handling will apply to parse message. * Added query processing support for the Bind message.	10 months ago
Rahim Kanji	344ba25777	Add Watchdog PROXYSQLTEST covering MySQL and PgSQL threads	11 months ago
Rahim Kanji	745449fcbc	Removed Client_Session.h and Client_Session.cpp	2 years ago
Rene Cannao	0481ab5936	Adding some timers for debugging purposes ... Adding Timers in Base_Thread to count the number of time and elapsed time of when Session::handler() and Connection::handler() are called. This feature is disabled by default. Timers are enabled only if in proxysql_macros.h the following is defined: #define ENABLE_TIMER true The default: #define ENABLE_TIMER false	2 years ago
René Cannaò	2d7806c325	Moved run_SetAllSession_ToProcess0 to Base_Thread	2 years ago
René Cannaò	483b8335d3	More functions moved to Base_Thread ... * move_session_to_idle_mysql_sessions * find_session_idx_in_mysql_sessions * ProcessAllMyDS_BeforePoll	2 years ago
René Cannaò	a881936cad	More functions moved to Base_Thread ... * ProcessAllMyDS_AfterPoll * read_one_byte_from_pipe * tune_timeout_for_myds_needs_pause * tune_timeout_for_session_needs_pause * configure_pollout * set_backend_to_be_skipped_if_frontend_is_slow	2 years ago
René Cannaò	db6cb93fb9	Moved ProcessAllSessions_SortingSessions to Base_Thread	2 years ago
René Cannaò	8389f76413	More functions moved to Base_Thread() ... * check_timing_out_session() * check_for_invalid_fd()	2 years ago
René Cannaò	c9326564c0	Minor changes to try to optimize compiling	2 years ago
René Cannaò	c903a45633	Second attempt to use Base_Thread ... Moved register_session() into Base_Thread. Moved Session_Regex() from Client_Session() to Base_Session() just to fix some compiler issue. Also removed some dead code from previous commit.	2 years ago
René Cannaò	0cfab2616f	First attempt to use Base_Thread and Base_Session	2 years ago