mirror of
https://github.com/mariadb-corporation/mariadb-columnstore-engine.git
synced 2025-05-10 02:01:04 +03:00
Part 1: As part of MCOL-3776 to address synchronization issue while accessing the fTimeZone member of the Func class, mutex locks were added to the accessor and mutator methods. However, this slows down processing of TIMESTAMP columns in PrimProc significantly as all threads across all concurrently running queries would serialize on the mutex. This is because PrimProc only has a single global object for the functor class (class derived from Func in utils/funcexp/functor.h) for a given function name. To fix this problem: (1) We remove the fTimeZone as a member of the Func derived classes (hence removing the mutexes) and instead use the fOperationType member of the FunctionColumn class to propagate the timezone values down to the individual functor processing functions such as FunctionColumn::getStrVal(), FunctionColumn::getIntVal(), etc. (2) To achieve (1), a timezone member is added to the execplan::CalpontSystemCatalog::ColType class. Part 2: Several functors in the Funcexp code call dataconvert::gmtSecToMySQLTime() and dataconvert::mySQLTimeToGmtSec() functions for conversion between seconds since unix epoch and broken-down representation. These functions in turn call the C library function localtime_r() which currently has a known bug of holding a global lock via a call to __tz_convert. This significantly reduces performance in multi-threaded applications where multiple threads concurrently call localtime_r(). More details on the bug: https://sourceware.org/bugzilla/show_bug.cgi?id=16145 This bug in localtime_r() caused processing of the Functors in PrimProc to slowdown significantly since a query execution causes Functors code to be processed in a multi-threaded manner. As a fix, we remove the calls to localtime_r() from gmtSecToMySQLTime() and mySQLTimeToGmtSec() by performing the timezone-to-offset conversion (done in dataconvert::timeZoneToOffset()) during the execution plan creation in the plugin. Note that localtime_r() is only called when the time_zone system variable is set to "SYSTEM". This fix also required changing the timezone type from a std::string to a long across the system.
345 lines
11 KiB
C++
345 lines
11 KiB
C++
/* Copyright (C) 2014 InfiniDB, Inc.
|
|
Copyright (C) 2016 MariaDB Corporation
|
|
Copyright (C) 2019 MariaDB Corporation
|
|
|
|
This program is free software; you can redistribute it and/or
|
|
modify it under the terms of the GNU General Public License
|
|
as published by the Free Software Foundation; version 2 of
|
|
the License.
|
|
|
|
This program is distributed in the hope that it will be useful,
|
|
but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
GNU General Public License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this program; if not, write to the Free Software
|
|
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
|
|
MA 02110-1301, USA. */
|
|
#pragma once
|
|
|
|
#include <my_config.h>
|
|
#include "idb_mysql.h"
|
|
#include "ha_mcs_sysvars.h"
|
|
#include "ha_maria.h"
|
|
|
|
extern handlerton* mcs_hton;
|
|
|
|
/** @brief
|
|
COLUMNSTORE_SHARE is a structure that will be shared among all open handlers.
|
|
This example implements the minimum of what you will probably need.
|
|
*/
|
|
typedef struct st_mcs_share
|
|
{
|
|
char* table_name;
|
|
uint32_t table_name_length, use_count;
|
|
pthread_mutex_t mutex;
|
|
THR_LOCK lock;
|
|
} COLUMNSTORE_SHARE;
|
|
|
|
/** @brief
|
|
Class definition for the storage engine
|
|
*/
|
|
class ha_mcs : public handler
|
|
{
|
|
THR_LOCK_DATA lock; ///< MySQL lock
|
|
COLUMNSTORE_SHARE* share; ///< Shared lock info
|
|
ulonglong int_table_flags;
|
|
// We are using a vector here to mimick the stack functionality
|
|
// using push_back() and pop_back()
|
|
// as apparently there is a linker error on the std::stack<COND*>::pop()
|
|
// call on Ubuntu18.
|
|
std::vector<COND*> condStack;
|
|
int m_lock_type;
|
|
long time_zone;
|
|
|
|
int impl_external_lock(THD* thd, TABLE* table, int lock_type);
|
|
int impl_rnd_init(TABLE* table, const std::vector<COND*>& condStack);
|
|
|
|
public:
|
|
ha_mcs(handlerton* hton, TABLE_SHARE* table_arg);
|
|
virtual ~ha_mcs()
|
|
{
|
|
}
|
|
|
|
/** @brief
|
|
The name that will be used for display purposes.
|
|
*/
|
|
const char* table_type() const override
|
|
{
|
|
return "ColumnStore";
|
|
}
|
|
|
|
/** @brief
|
|
The name of the index type that will be used for display.
|
|
Don't implement this method unless you really have indexes.
|
|
*/
|
|
// const char *index_type(uint32_t inx) { return "HASH"; }
|
|
|
|
/** @brief
|
|
The file extensions.
|
|
*/
|
|
const char** bas_ext() const;
|
|
|
|
/** @brief
|
|
This is a list of flags that indicate what functionality the storage engine
|
|
implements. The current table flags are documented in handler.h
|
|
*/
|
|
ulonglong table_flags() const override
|
|
{
|
|
return int_table_flags;
|
|
}
|
|
|
|
/** @brief
|
|
This is a bitmap of flags that indicates how the storage engine
|
|
implements indexes. The current index flags are documented in
|
|
handler.h. If you do not implement indexes, just return zero here.
|
|
|
|
@details
|
|
part is the key part to check. First key part is 0.
|
|
If all_parts is set, MySQL wants to know the flags for the combined
|
|
index, up to and including 'part'.
|
|
*/
|
|
ulong index_flags(uint32_t inx, uint32_t part, bool all_parts) const override
|
|
{
|
|
return 0;
|
|
}
|
|
|
|
/** @brief
|
|
unireg.cc will call max_supported_record_length(), max_supported_keys(),
|
|
max_supported_key_parts(), uint32_t max_supported_key_length()
|
|
to make sure that the storage engine can handle the data it is about to
|
|
send. Return *real* limits of your storage engine here; MySQL will do
|
|
min(your_limits, MySQL_limits) automatically.
|
|
*/
|
|
uint32_t max_supported_record_length() const override
|
|
{
|
|
return HA_MAX_REC_LENGTH;
|
|
}
|
|
|
|
/** @brief
|
|
Called in test_quick_select to determine if indexes should be used.
|
|
*/
|
|
virtual double scan_time() override
|
|
{
|
|
return (double)(stats.records + stats.deleted) / 20.0 + 10;
|
|
}
|
|
|
|
/** @brief
|
|
Analyze table command.
|
|
*/
|
|
int analyze(THD* thd, HA_CHECK_OPT* check_opt) override;
|
|
|
|
/*
|
|
Everything below are methods that we implement in ha_example.cc.
|
|
|
|
Most of these methods are not obligatory, skip them and
|
|
MySQL will treat them as not implemented
|
|
*/
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc; it's a required method.
|
|
*/
|
|
int open(const char* name, int mode, uint32_t test_if_locked) override; // required
|
|
|
|
// MCOL-4282 This function is called by open_tables in sql_base.cc.
|
|
// We mutate the optimizer flags here for prepared statements as this
|
|
// handler function is called before JOIN::prepare, and we need to
|
|
// disable the default optimizer flags before JOIN::prepare (which is
|
|
// called during "PREPARE stmt FROM ..." SQL) is called.
|
|
// Sequence of SQL statements that will lead to this execution path
|
|
// for prepared statements:
|
|
// CREATE TABLE t1 (a int, b int) engine=columnstore;
|
|
// INSERT INTO t1 VALUES (1, 2), (2, 4), (3, 1);
|
|
// PREPARE stmt1 FROM "SELECT * FROM t1";
|
|
// EXECUTE stmt1;
|
|
|
|
int discover_check_version() override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc; it's a required method.
|
|
*/
|
|
int close(void) override; // required
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int write_row(const uchar* buf) override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
void start_bulk_insert(ha_rows rows, uint flags = 0) override;
|
|
void start_bulk_insert_from_cache(ha_rows rows, uint flags = 0);
|
|
|
|
/**@bug 2461 - Overloaded end_bulk_insert. MariaDB uses the abort bool, mysql does not. */
|
|
int end_bulk_insert() override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int update_row(const uchar* old_data, const uchar* new_data) override;
|
|
int direct_update_rows_init(List<Item>* update_fields) override;
|
|
int direct_update_rows(ha_rows* update_rows);
|
|
int direct_update_rows(ha_rows* update_rows, ha_rows* found_rows) override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int delete_row(const uchar* buf) override;
|
|
int direct_delete_rows_init() override;
|
|
int direct_delete_rows(ha_rows* deleted_rows) override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int index_read_map(uchar* buf, const uchar* key, key_part_map keypart_map,
|
|
enum ha_rkey_function find_flag) override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int index_next(uchar* buf) override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int index_prev(uchar* buf) override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int index_first(uchar* buf) override;
|
|
|
|
/** @brief
|
|
We implement this in ha_example.cc. It's not an obligatory method;
|
|
skip it and and MySQL will treat it as not implemented.
|
|
*/
|
|
int index_last(uchar* buf) override;
|
|
|
|
/** @brief
|
|
Unlike index_init(), rnd_init() can be called two consecutive times
|
|
without rnd_end() in between (it only makes sense if scan=1). In this
|
|
case, the second call should prepare for the new table scan (e.g if
|
|
rnd_init() allocates the cursor, the second call should position the
|
|
cursor to the start of the table; no need to deallocate and allocate
|
|
it again. This is a required method.
|
|
*/
|
|
int rnd_init(bool scan) override; // required
|
|
int rnd_end() override;
|
|
int rnd_next(uchar* buf) override; ///< required
|
|
int rnd_pos(uchar* buf, uchar* pos) override; ///< required
|
|
int reset() override;
|
|
void position(const uchar* record) override; ///< required
|
|
int info(uint32_t) override; ///< required
|
|
int extra(enum ha_extra_function operation) override;
|
|
int external_lock(THD* thd, int lock_type) override; ///< required
|
|
int delete_all_rows(void) override;
|
|
ha_rows records_in_range(uint32_t inx, const key_range* min_key, const key_range* max_key,
|
|
page_range* res) override;
|
|
int delete_table(const char* from) override;
|
|
int rename_table(const char* from, const char* to) override;
|
|
int create(const char* name, TABLE* form,
|
|
HA_CREATE_INFO* create_info) override; ///< required
|
|
|
|
THR_LOCK_DATA** store_lock(THD* thd, THR_LOCK_DATA** to,
|
|
enum thr_lock_type lock_type) override; ///< required
|
|
const COND* cond_push(const COND* cond) override;
|
|
void cond_pop() override;
|
|
uint8 table_cache_type() override
|
|
{
|
|
return HA_CACHE_TBL_NOCACHE;
|
|
}
|
|
|
|
int repair(THD* thd, HA_CHECK_OPT* check_opt) override;
|
|
bool is_crashed() const override;
|
|
|
|
bool isReadOnly() const
|
|
{
|
|
return m_lock_type == F_RDLCK;
|
|
}
|
|
};
|
|
|
|
class ha_mcs_cache_share
|
|
{
|
|
ha_mcs_cache_share* next; /* Next open share */
|
|
const char* name;
|
|
uint open_count;
|
|
|
|
public:
|
|
ulonglong cached_rows;
|
|
THR_LOCK org_lock;
|
|
friend ha_mcs_cache_share* find_cache_share(const char* name, ulonglong cached_rows);
|
|
void close();
|
|
};
|
|
|
|
class ha_mcs_cache : public ha_mcs
|
|
{
|
|
typedef ha_mcs parent;
|
|
int original_lock_type;
|
|
bool insert_command, cache_locked;
|
|
enum_sql_command sql_command;
|
|
|
|
// True if this handler belongs to either calpontsys.systable or
|
|
// calpontsys.syscolumn system catalog tables
|
|
bool isSysCatTable;
|
|
|
|
// True if the ColumnStore table is not cached (i.e. when the table
|
|
// was created with columnstore_cache_inserts=OFF).
|
|
bool isCacheDisabled;
|
|
|
|
bool isCacheEnabled() const
|
|
{
|
|
return (get_cache_inserts(current_thd) && !isSysCatTable && !isCacheDisabled);
|
|
}
|
|
|
|
public:
|
|
uint lock_counter;
|
|
ha_maria* cache_handler;
|
|
ha_mcs_cache_share* share;
|
|
|
|
ha_mcs_cache(handlerton* hton, TABLE_SHARE* table_arg, MEM_ROOT* mem_root);
|
|
~ha_mcs_cache();
|
|
|
|
/*
|
|
The following functions duplicates calls to derived handler and
|
|
cache handler
|
|
*/
|
|
|
|
int create(const char* name, TABLE* table_arg, HA_CREATE_INFO* ha_create_info);
|
|
int open(const char* name, int mode, uint open_flags);
|
|
int delete_table(const char* name);
|
|
int rename_table(const char* from, const char* to);
|
|
int delete_all_rows(void);
|
|
int close(void);
|
|
|
|
uint lock_count(void) const;
|
|
THR_LOCK_DATA** store_lock(THD* thd, THR_LOCK_DATA** to, enum thr_lock_type lock_type);
|
|
int external_lock(THD* thd, int lock_type);
|
|
int repair(THD* thd, HA_CHECK_OPT* check_opt);
|
|
bool is_crashed() const;
|
|
|
|
/*
|
|
Write row uses cache_handler, for normal inserts, otherwise derived
|
|
handler
|
|
*/
|
|
int write_row(const uchar* buf);
|
|
void start_bulk_insert(ha_rows rows, uint flags);
|
|
int end_bulk_insert();
|
|
|
|
/* Cache functions */
|
|
void free_locks();
|
|
ha_rows num_rows_cached();
|
|
int flush_insert_cache();
|
|
friend my_bool get_status_and_flush_cache(void* param, my_bool concurrent_insert);
|
|
friend my_bool cache_start_trans(void* param);
|
|
};
|