#include <ddl0bulk.h>
◆ Btree_loads
◆ Thread_ctxs
◆ Loader()
ddl_bulk::Loader::Loader |
( |
size_t |
num_threads | ) |
|
|
inline |
Loader context constructor.
- Parameters
-
[in] | num_threads | Number of threads to use for bulk loading |
◆ begin()
Prepare bulk loading by multiple threads.
- Parameters
-
[in] | prebuilt | prebuilt structures from innodb table handler |
[in] | data_size | total data size to load in bytes |
[in] | memory | memory to be used from buffer pool |
- Returns
- innodb error code
◆ end()
Finish bulk load operation, combining the sub-trees produced by concurrent threads.
- Parameters
-
[in] | prebuilt | prebuilt structures from innodb table handler |
[in] | is_error | true if called for cleanup and rollback after an error |
- Returns
- innodb error code
◆ get_error()
dberr_t Loader::get_error |
( |
| ) |
const |
|
inline |
◆ get_error_code()
int Loader::get_error_code |
( |
| ) |
const |
|
inline |
Get the client error code (e.g.
ER_LOAD_BULK_DATA_UNSORTED).
- Returns
- the client error code.
◆ get_error_string()
std::string Loader::get_error_string |
( |
| ) |
const |
|
inline |
◆ get_index_name()
const char * ddl_bulk::Loader::get_index_name |
( |
| ) |
const |
|
inline |
- Returns
- index name where the data is being loaded.
◆ get_queue_size()
void Loader::get_queue_size |
( |
size_t |
memory, |
|
|
size_t & |
flush_queue_size, |
|
|
bool & |
allocate_in_pages |
|
) |
| const |
|
private |
Calculate the flush queue size to be used based on the available memory.
- Parameters
-
[in] | memory | total buffer pool memory to use |
[out] | flush_queue_size | calculated queue size |
[out] | allocate_in_pages | true if need to allocate in pages false if need to allocate in extents |
◆ get_table_name()
const char * ddl_bulk::Loader::get_table_name |
( |
| ) |
const |
|
inline |
- Returns
- table name where the data is being loaded.
◆ load()
Load rows to a sub-tree by a thread.
Called concurrently by multiple execution threads.
- Parameters
-
[in] | prebuilt | prebuilt structures from innodb table handler |
[in] | thread_index | true if called for cleanup and rollback after an error |
[in] | rows | rows to be loaded to the cluster index sub-tree |
[in] | wait_cbk | Stat callbacks |
- Returns
- innodb error code
◆ merge_subtrees()
Merge the sub-trees to build the cluster index.
- Parameters
-
[in] | prebuilt | prebuilt structures from innodb table handler |
- Returns
- innodb error code.
◆ m_ctxs
All thread specific data.
◆ m_extent_allocator
Allocator to extend tablespace and allocate extents.
◆ m_num_threads
size_t ddl_bulk::Loader::m_num_threads {} |
|
private |
Number of threads for bulk loading.
◆ m_sub_tree_loads
Sub-tree loading contexts.
◆ m_table
Innodb dictionary table object.
The documentation for this class was generated from the following files: