InnoDB Plugin
1.0
|
#include <log0log.h>
Data Fields | |
byte | pad [64] |
lsn_t | lsn |
ulint | buf_free |
ib_mutex_t | mutex |
ib_mutex_t | log_flush_order_mutex |
byte * | buf_ptr |
byte * | buf |
ulint | buf_size |
ulint | max_buf_free |
ibool | check_flush_or_checkpoint |
ut_list_base< log_group_t > | log_groups |
ulint | buf_next_to_write |
volatile bool | is_extending |
lsn_t | written_to_some_lsn |
lsn_t | written_to_all_lsn |
lsn_t | write_lsn |
ulint | write_end_offset |
lsn_t | current_flush_lsn |
lsn_t | flushed_to_disk_lsn |
ulint | n_pending_writes |
os_event_t | no_flush_event |
ibool | one_flushed |
os_event_t | one_flushed_event |
ulint | n_log_ios |
ulint | n_log_ios_old |
time_t | last_printout_time |
lsn_t | log_group_capacity |
lsn_t | max_modified_age_async |
lsn_t | max_modified_age_sync |
lsn_t | max_checkpoint_age_async |
lsn_t | max_checkpoint_age |
ib_uint64_t | next_checkpoint_no |
lsn_t | last_checkpoint_lsn |
lsn_t | next_checkpoint_lsn |
ulint | n_pending_checkpoint_writes |
rw_lock_t | checkpoint_lock |
byte * | checkpoint_buf_ptr |
byte * | checkpoint_buf |
Redo log buffer
byte* log_t::buf |
log buffer
ulint log_t::buf_free |
first free offset within the log buffer
ulint log_t::buf_next_to_write |
The fields involved in the log buffer flush first offset in the log buffer where the byte content may not exist written to file, e.g., the start offset of a log record catenated later; this is advanced when a flush operation is completed to all the log groups
ulint log_t::buf_size |
log buffer size in bytes
ibool log_t::check_flush_or_checkpoint |
this is set to TRUE when there may be need to flush the log buffer, or preflush buffer pool pages, or make a checkpoint; this MUST be TRUE when lsn - last_checkpoint_lsn > max_checkpoint_age; this flag is peeked at by log_free_check(), which does not reserve the log mutex
byte* log_t::checkpoint_buf |
checkpoint header is read to this buffer
rw_lock_t log_t::checkpoint_lock |
this latch is x-locked when a checkpoint write is running; a thread should wait for this without owning the log mutex
lsn_t log_t::current_flush_lsn |
end lsn for the current running write + flush operation
lsn_t log_t::flushed_to_disk_lsn |
how far we have written the log AND flushed to disk
volatile bool log_t::is_extending |
this is set to true during extend the log buffer size
lsn_t log_t::last_checkpoint_lsn |
latest checkpoint lsn
time_t log_t::last_printout_time |
when log_print was last time called
ib_mutex_t log_t::log_flush_order_mutex |
mutex to serialize access to the flush list when we are putting dirty blocks in the list. The idea behind this mutex is to be able to release log_sys->mutex during mtr_commit and still ensure that insertions in the flush_list happen in the LSN order.
lsn_t log_t::log_group_capacity |
Fields involved in checkpoints capacity of the log group; if the checkpoint age exceeds this, it is a serious error because it is possible we will then overwrite log and spoil crash recovery
ut_list_base< log_group_t > log_t::log_groups |
log groups
lsn_t log_t::lsn |
log sequence number
ulint log_t::max_buf_free |
recommended maximum value of buf_free, after which the buffer is flushed
lsn_t log_t::max_checkpoint_age |
this is the maximum allowed value for lsn - last_checkpoint_lsn when a new query step is started
lsn_t log_t::max_checkpoint_age_async |
when this checkpoint age is exceeded we start an asynchronous writing of a new checkpoint
lsn_t log_t::max_modified_age_async |
when this recommended value for lsn - buf_pool_get_oldest_modification() is exceeded, we start an asynchronous preflush of pool pages
lsn_t log_t::max_modified_age_sync |
when this recommended value for lsn - buf_pool_get_oldest_modification() is exceeded, we start a synchronous preflush of pool pages
ib_mutex_t log_t::mutex |
mutex protecting the log
ulint log_t::n_log_ios |
number of log i/os initiated thus far
ulint log_t::n_log_ios_old |
number of log i/o's at the previous printout
ulint log_t::n_pending_checkpoint_writes |
number of currently pending checkpoint writes
ulint log_t::n_pending_writes |
number of currently pending flushes or writes
lsn_t log_t::next_checkpoint_lsn |
next checkpoint lsn
ib_uint64_t log_t::next_checkpoint_no |
next checkpoint number
os_event_t log_t::no_flush_event |
this event is in the reset state when a flush or a write is running; a thread should wait for this without owning the log mutex, but NOTE that to set or reset this event, the thread MUST own the log mutex!
ibool log_t::one_flushed |
during a flush, this is first FALSE and becomes TRUE when one log group has been written or flushed
os_event_t log_t::one_flushed_event |
this event is reset when the flush or write has not yet completed for any log group; e.g., this means that a transaction has been committed when this is set; a thread should wait for this without owning the log mutex, but NOTE that to set or reset this event, the thread MUST own the log mutex!
byte log_t::pad[64] |
padding to prevent other memory update hotspots from residing on the same memory cache line
ulint log_t::write_end_offset |
the data in buffer has been written up to this offset when the current write ends: this field will then be copied to buf_next_to_write
lsn_t log_t::write_lsn |
end lsn for the current running write
lsn_t log_t::written_to_all_lsn |
first log sequence number not yet written to some log group; for this to be advanced, it is enough that the write i/o has been completed for all log groups. Note that since InnoDB currently has only one log group therefore this value is redundant. Also it is possible that this value falls behind the flushed_to_disk_lsn transiently. It is appropriate to use either flushed_to_disk_lsn or write_lsn which are always up-to-date and accurate.
lsn_t log_t::written_to_some_lsn |
first log sequence number not yet written to any log group; for this to be advanced, it is enough that the write i/o has been completed for any one log group