packages/mysql5
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
mysql5/mysql-5.7.27/sql/rpl_rli_pdb.cc

2756 lines
85 KiB
C++
Raw Blame History

/* Copyright (c) 2011, 2018, Oracle and/or its affiliates. All rights reserved.
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 St, Fifth Floor, Boston, MA 02110-1301 USA */
#include "debug_sync.h"
#include "rpl_rli_pdb.h"
#include "log.h" // sql_print_error
#include "rpl_slave_commit_order_manager.h" // Commit_order_manager
#include "rpl_msr.h" // for channel_map
#include "pfs_file_provider.h"
#include "mysql/psi/mysql_file.h"
#ifndef DBUG_OFF
ulong w_rr= 0;
uint mts_debug_concurrent_access= 0;
#endif
#define HASH_DYNAMIC_INIT 4
using std::min;
using std::max;
/**
This function is called by both coordinator and workers.
Upon receiving the STOP command, the workers will identify a
maximum group index already executed (or under execution).
All groups whose index are below or equal to the maximum
group index will be applied by the workers before stopping.
The workers with groups above the maximum group index will
exit without applying these groups by setting their running
status to "STOP_ACCEPTED".
@param worker a pointer to the waiting Worker struct
@param job_item a pointer to struct carrying a reference to an event
@return true if STOP command gets accepted otherwise false is returned.
*/
bool handle_slave_worker_stop(Slave_worker *worker,
Slave_job_item *job_item)
{
ulonglong group_index= 0;
Relay_log_info *rli= worker->c_rli;
mysql_mutex_lock(&rli->exit_count_lock);
/*
First, W calculates a group-"at-hands" index which is
either the currently read ev group index, or the last executed
group's one when the queue is empty.
*/
group_index= (job_item->data)?
rli->gaq->get_job_group(job_item->data->mts_group_idx)->total_seqno:
worker->last_groups_assigned_index;
/*
The max updated index is being updated as long as
exit_counter permits. That's stopped with the final W's
increment of it.
*/
if (!worker->exit_incremented)
{
if (rli->exit_counter < rli->slave_parallel_workers)
rli->max_updated_index = max(rli->max_updated_index, group_index);
++rli->exit_counter;
worker->exit_incremented= true;
DBUG_ASSERT(!is_mts_worker(current_thd));
}
#ifndef DBUG_OFF
else
DBUG_ASSERT(is_mts_worker(current_thd));
#endif
/*
Now let's decide about the deferred exit to consider
the empty queue and the counter value reached
slave_parallel_workers.
*/
if (!job_item->data)
{
worker->running_status= Slave_worker::STOP_ACCEPTED;
mysql_cond_signal(&worker->jobs_cond);
mysql_mutex_unlock(&rli->exit_count_lock);
return(true);
}
else if (rli->exit_counter == rli->slave_parallel_workers)
{
//over steppers should exit with accepting STOP
if (group_index > rli->max_updated_index)
{
worker->running_status= Slave_worker::STOP_ACCEPTED;
mysql_cond_signal(&worker->jobs_cond);
mysql_mutex_unlock(&rli->exit_count_lock);
return(true);
}
}
mysql_mutex_unlock(&rli->exit_count_lock);
return(false);
}
/**
This function is called by both coordinator and workers.
Both coordinator and workers contribute to max_updated_index.
@param worker a pointer to the waiting Worker struct
@param job_item a pointer to struct carrying a reference to an event
@return true if STOP command gets accepted otherwise false is returned.
*/
bool set_max_updated_index_on_stop(Slave_worker *worker,
Slave_job_item *job_item)
{
head_queue(&worker->jobs, job_item);
if (worker->running_status == Slave_worker::STOP)
{
if (handle_slave_worker_stop(worker, job_item))
return true;
}
return false;
}
/*
Please every time you add a new field to the worker slave info, update
what follows. For now, this is just used to get the number of fields.
*/
const char *info_slave_worker_fields []=
{
"id",
/*
These positions identify what has been executed. Notice that they are
redudant and only the group_master_log_name and group_master_log_pos
are really necessary. However, the additional information is kept to
ease debugging.
*/
"group_relay_log_name",
"group_relay_log_pos",
"group_master_log_name",
"group_master_log_pos",
/*
These positions identify what a worker knew about the coordinator at
the time a job was assigned. Notice that they are redudant and are
kept to ease debugging.
*/
"checkpoint_relay_log_name",
"checkpoint_relay_log_pos",
"checkpoint_master_log_name",
"checkpoint_master_log_pos",
/*
Identify the greatest job, i.e. group, processed by a worker.
*/
"checkpoint_seqno",
/*
Maximum number of jobs that can be assigned to a worker. This
information is necessary to read the next entry.
*/
"checkpoint_group_size",
/*
Bitmap used to identify what jobs were processed by a worker.
*/
"checkpoint_group_bitmap",
/*
Channel on which this workers are acting
*/
"channel_name"
};
/*
Number of records in the mts partition hash below which
entries with zero usage are tolerated so could be quickly
recycled.
*/
const ulong mts_partition_hash_soft_max= 16;
/*
index value of some outstanding slots of info_slave_worker_fields
*/
enum {
LINE_FOR_CHANNEL= 12,
};
const uint info_slave_worker_table_pk_field_indexes []=
{
LINE_FOR_CHANNEL,
0,
};
Slave_worker::Slave_worker(Relay_log_info *rli
#ifdef HAVE_PSI_INTERFACE
,PSI_mutex_key *param_key_info_run_lock,
PSI_mutex_key *param_key_info_data_lock,
PSI_mutex_key *param_key_info_sleep_lock,
PSI_mutex_key *param_key_info_thd_lock,
PSI_mutex_key *param_key_info_data_cond,
PSI_mutex_key *param_key_info_start_cond,
PSI_mutex_key *param_key_info_stop_cond,
PSI_mutex_key *param_key_info_sleep_cond
#endif
, uint param_id, const char *param_channel
)
: Relay_log_info(FALSE
#ifdef HAVE_PSI_INTERFACE
,param_key_info_run_lock, param_key_info_data_lock,
param_key_info_sleep_lock, param_key_info_thd_lock,
param_key_info_data_cond, param_key_info_start_cond,
param_key_info_stop_cond, param_key_info_sleep_cond
#endif
, param_id + 1, param_channel, true
),
c_rli(rli),
curr_group_exec_parts(key_memory_db_worker_hash_entry),
id(param_id),
checkpoint_relay_log_pos(0), checkpoint_master_log_pos(0),
checkpoint_seqno(0), running_status(NOT_RUNNING), exit_incremented(false)
{
/*
In the future, it would be great if we use only one identifier.
So when factoring out this code, please, consider this.
*/
DBUG_ASSERT(internal_id == id + 1);
checkpoint_relay_log_name[0]= 0;
checkpoint_master_log_name[0]= 0;
mysql_mutex_init(key_mutex_slave_parallel_worker, &jobs_lock,
MY_MUTEX_INIT_FAST);
mysql_cond_init(key_cond_slave_parallel_worker, &jobs_cond);
mysql_cond_init(key_cond_mts_gaq, &logical_clock_cond);
}
Slave_worker::~Slave_worker()
{
end_info();
if (jobs.inited_queue)
{
DBUG_ASSERT(jobs.m_Q.size() == jobs.size);
jobs.m_Q.clear();
}
mysql_mutex_destroy(&jobs_lock);
mysql_cond_destroy(&jobs_cond);
mysql_cond_destroy(&logical_clock_cond);
mysql_mutex_lock(&info_thd_lock);
info_thd= NULL;
mysql_mutex_unlock(&info_thd_lock);
set_rli_description_event(NULL);
}
/**
Method is executed by Coordinator at Worker startup time to initialize
members parly with values supplied by Coordinator through rli.
@param rli Coordinator's Relay_log_info pointer
@param i identifier of the Worker
@return 0 success
non-zero failure
*/
int Slave_worker::init_worker(Relay_log_info * rli, ulong i)
{
DBUG_ENTER("Slave_worker::init_worker");
DBUG_ASSERT(!rli->info_thd->is_error());
Slave_job_item empty= Slave_job_item();
c_rli= rli;
set_commit_order_manager(c_rli->get_commit_order_manager());
if (rli_init_info(false) ||
DBUG_EVALUATE_IF("inject_init_worker_init_info_fault", true, false))
DBUG_RETURN(1);
id= i;
curr_group_exec_parts.clear();
relay_log_change_notified= FALSE; // the 1st group to contain relaylog name
checkpoint_notified= FALSE; // the same as above
master_log_change_notified= false;// W learns master log during 1st group exec
fd_change_notified= false; // W is to learn master FD version same as above
server_version= version_product(rli->slave_version_split);
bitmap_shifted= 0;
workers= c_rli->workers; // shallow copying is sufficient
wq_empty_waits= wq_size_waits_cnt= groups_done= events_done= curr_jobs= 0;
usage_partition= 0;
end_group_sets_max_dbs= false;
gaq_index= last_group_done_index= c_rli->gaq->size; // out of range
last_groups_assigned_index=0;
DBUG_ASSERT(!jobs.inited_queue);
jobs.avail= 0;
jobs.len= 0;
jobs.overfill= FALSE; // todo: move into Slave_jobs_queue constructor
jobs.waited_overfill= 0;
jobs.entry= jobs.size= c_rli->mts_slave_worker_queue_len_max;
jobs.inited_queue= true;
curr_group_seen_begin= curr_group_seen_gtid= false;
#ifndef DBUG_OFF
curr_group_seen_sequence_number= false;
#endif
jobs.m_Q.resize(jobs.size, empty);
DBUG_ASSERT(jobs.m_Q.size() == jobs.size);
wq_overrun_cnt= excess_cnt= 0;
underrun_level= (ulong) ((rli->mts_worker_underrun_level * jobs.size) / 100.0);
// overrun level is symmetric to underrun (as underrun to the full queue)
overrun_level= jobs.size - underrun_level;
/* create mts submode for each of the the workers. */
current_mts_submode=
(rli->channel_mts_submode == MTS_PARALLEL_TYPE_DB_NAME)?
(Mts_submode*) new Mts_submode_database():
(Mts_submode*) new Mts_submode_logical_clock();
//workers and coordinator must be of the same type
DBUG_ASSERT(rli->current_mts_submode->get_type() ==
current_mts_submode->get_type());
m_order_commit_deadlock= false;
DBUG_RETURN(0);
}
/**
A part of Slave worker iitializer that provides a
minimum context for MTS recovery.
@param is_gaps_collecting_phase
clarifies what state the caller
executes this method from. When it's @c true
that is @c mts_recovery_groups() and Worker should
restore the last session time info which is processed
to collect gaps that is not executed transactions (groups).
Such recovery Slave_worker intance is destroyed at the end of
@c mts_recovery_groups().
Whet it's @c false Slave_worker is initialized for the run time
nad should not read the last session time stale info.
Its info will be ultimately reset once all gaps are executed
to finish off recovery.
@return 0 on success, non-zero for a failure
*/
int Slave_worker::rli_init_info(bool is_gaps_collecting_phase)
{
enum_return_check return_check= ERROR_CHECKING_REPOSITORY;
DBUG_ENTER("Slave_worker::rli_init_info");
if (inited)
DBUG_RETURN(0);
/*
Worker bitmap size depends on recovery mode.
If it is gaps collecting the bitmaps must be capable to accept
up to MTS_MAX_BITS_IN_GROUP of bits.
*/
size_t num_bits= is_gaps_collecting_phase ?
MTS_MAX_BITS_IN_GROUP : c_rli->checkpoint_group;
/*
This checks if the repository was created before and thus there
will be values to be read. Please, do not move this call after
the handler->init_info().
*/
return_check= check_info();
if (return_check == ERROR_CHECKING_REPOSITORY ||
(return_check == REPOSITORY_DOES_NOT_EXIST && is_gaps_collecting_phase))
goto err;
if (handler->init_info())
goto err;
bitmap_init(&group_executed, NULL, num_bits, FALSE);
bitmap_init(&group_shifted, NULL, num_bits, FALSE);
if (is_gaps_collecting_phase &&
(DBUG_EVALUATE_IF("mts_slave_worker_init_at_gaps_fails", true, false) ||
read_info(handler)))
{
bitmap_free(&group_executed);
bitmap_free(&group_shifted);
goto err;
}
inited= 1;
DBUG_RETURN(0);
err:
// todo: handler->end_info(uidx, nidx);
inited= 0;
sql_print_error("Error reading slave worker configuration");
DBUG_RETURN(1);
}
void Slave_worker::end_info()
{
DBUG_ENTER("Slave_worker::end_info");
if (!inited)
DBUG_VOID_RETURN;
if (handler)
handler->end_info();
if (inited)
{
bitmap_free(&group_executed);
bitmap_free(&group_shifted);
}
inited = 0;
DBUG_VOID_RETURN;
}
int Slave_worker::flush_info(const bool force)
{
DBUG_ENTER("Slave_worker::flush_info");
if (!inited)
DBUG_RETURN(0);
/*
We update the sync_period at this point because only here we
now that we are handling a Slave_worker. This needs to be
update every time we call flush because the option may be
dinamically set.
*/
handler->set_sync_period(sync_relayloginfo_period);
if (write_info(handler))
goto err;
if (handler->flush_info(force))
goto err;
DBUG_RETURN(0);
err:
sql_print_error("Error writing slave worker configuration");
DBUG_RETURN(1);
}
bool Slave_worker::read_info(Rpl_info_handler *from)
{
DBUG_ENTER("Slave_worker::read_info");
ulong temp_group_relay_log_pos= 0;
ulong temp_group_master_log_pos= 0;
ulong temp_checkpoint_relay_log_pos= 0;
ulong temp_checkpoint_master_log_pos= 0;
ulong temp_checkpoint_seqno= 0;
ulong nbytes= 0;
uchar *buffer= (uchar *) group_executed.bitmap;
int temp_internal_id= 0;
if (from->prepare_info_for_read())
DBUG_RETURN(TRUE);
if (from->get_info(&temp_internal_id, 0) ||
from->get_info(group_relay_log_name,
sizeof(group_relay_log_name),
(char *) "") ||
from->get_info(&temp_group_relay_log_pos,
0UL) ||
from->get_info(group_master_log_name,
sizeof(group_master_log_name),
(char *) "") ||
from->get_info(&temp_group_master_log_pos,
0UL) ||
from->get_info(checkpoint_relay_log_name,
sizeof(checkpoint_relay_log_name),
(char *) "") ||
from->get_info(&temp_checkpoint_relay_log_pos,
0UL) ||
from->get_info(checkpoint_master_log_name,
sizeof(checkpoint_master_log_name),
(char *) "") ||
from->get_info(&temp_checkpoint_master_log_pos,
0UL) ||
from->get_info(&temp_checkpoint_seqno,
0UL) ||
from->get_info(&nbytes, 0UL) ||
from->get_info(buffer, (size_t) nbytes,
(uchar *) 0) ||
/* default is empty string */
from->get_info(channel, sizeof(channel),(char*)""))
DBUG_RETURN(TRUE);
DBUG_ASSERT(nbytes <= no_bytes_in_map(&group_executed));
internal_id=(uint) temp_internal_id;
group_relay_log_pos= temp_group_relay_log_pos;
group_master_log_pos= temp_group_master_log_pos;
checkpoint_relay_log_pos= temp_checkpoint_relay_log_pos;
checkpoint_master_log_pos= temp_checkpoint_master_log_pos;
checkpoint_seqno= temp_checkpoint_seqno;
DBUG_RETURN(FALSE);
}
/*
This function is used to make a copy of the worker object before we
destroy it while STOP SLAVE. This new object is then used to report the
worker status until next START SLAVE following which the new worker objetcs
will be used.
*/
void Slave_worker::copy_values_for_PFS(ulong worker_id,
en_running_state thd_running_status,
THD *worker_thd,
const Error &last_error,
const Gtid_specification &gtid)
{
id= worker_id;
running_status= thd_running_status;
info_thd= worker_thd;
m_last_error= last_error;
currently_executing_gtid= gtid;
}
bool Slave_worker::set_info_search_keys(Rpl_info_handler *to)
{
DBUG_ENTER("Slave_worker::set_info_search_keys");
/* primary keys are Id and channel_name */
if(to->set_info(0, (int)internal_id ) || to->set_info(LINE_FOR_CHANNEL, channel))
DBUG_RETURN(TRUE);
DBUG_RETURN(FALSE);
}
bool Slave_worker::write_info(Rpl_info_handler *to)
{
DBUG_ENTER("Master_info::write_info");
ulong nbytes= (ulong) no_bytes_in_map(&group_executed);
uchar *buffer= (uchar*) group_executed.bitmap;
DBUG_ASSERT(nbytes <= (c_rli->checkpoint_group + 7) / 8);
if (to->prepare_info_for_write() ||
to->set_info((int) internal_id) ||
to->set_info(group_relay_log_name) ||
to->set_info((ulong) group_relay_log_pos) ||
to->set_info(group_master_log_name) ||
to->set_info((ulong) group_master_log_pos) ||
to->set_info(checkpoint_relay_log_name) ||
to->set_info((ulong) checkpoint_relay_log_pos) ||
to->set_info(checkpoint_master_log_name) ||
to->set_info((ulong) checkpoint_master_log_pos) ||
to->set_info((ulong) checkpoint_seqno) ||
to->set_info(nbytes) ||
to->set_info(buffer, (size_t) nbytes)||
to->set_info(channel))
DBUG_RETURN(TRUE);
DBUG_RETURN(FALSE);
}
/**
Clean up a part of Worker info table that is regarded in
in gaps collecting at recovery.
This worker won't contribute to recovery bitmap at future
slave restart (see @c mts_recovery_groups).
@retrun FALSE as success TRUE as failure
*/
bool Slave_worker::reset_recovery_info()
{
DBUG_ENTER("Slave_worker::reset_recovery_info");
set_group_master_log_name("");
set_group_master_log_pos(0);
DBUG_RETURN(flush_info(true));
}
size_t Slave_worker::get_number_worker_fields()
{
return sizeof(info_slave_worker_fields)/sizeof(info_slave_worker_fields[0]);
}
const char* Slave_worker::get_master_log_name()
{
Slave_job_group* ptr_g= c_rli->gaq->get_job_group(gaq_index);
return (ptr_g->checkpoint_log_name != NULL) ?
ptr_g->checkpoint_log_name : checkpoint_master_log_name;
}
bool Slave_worker::commit_positions(Log_event *ev, Slave_job_group* ptr_g, bool force)
{
DBUG_ENTER("Slave_worker::checkpoint_positions");
/*
Initial value of checkpoint_master_log_name is learned from
group_master_log_name. The latter can be passed to Worker
at rare event of master binlog rotation.
This initialization is needed to provide to Worker info
on physical coordiates during execution of the very first group
after a rotation.
*/
if (ptr_g->group_master_log_name != NULL)
{
strmake(group_master_log_name, ptr_g->group_master_log_name,
sizeof(group_master_log_name) - 1);
my_free(ptr_g->group_master_log_name);
ptr_g->group_master_log_name= NULL;
strmake(checkpoint_master_log_name, group_master_log_name,
sizeof(checkpoint_master_log_name) - 1);
}
if (ptr_g->checkpoint_log_name != NULL)
{
strmake(checkpoint_relay_log_name, ptr_g->checkpoint_relay_log_name,
sizeof(checkpoint_relay_log_name) - 1);
checkpoint_relay_log_pos= ptr_g->checkpoint_relay_log_pos;
strmake(checkpoint_master_log_name, ptr_g->checkpoint_log_name,
sizeof(checkpoint_master_log_name) - 1);
checkpoint_master_log_pos= ptr_g->checkpoint_log_pos;
my_free(ptr_g->checkpoint_log_name);
ptr_g->checkpoint_log_name= NULL;
my_free(ptr_g->checkpoint_relay_log_name);
ptr_g->checkpoint_relay_log_name= NULL;
bitmap_copy(&group_shifted, &group_executed);
bitmap_clear_all(&group_executed);
for (uint pos= ptr_g->shifted; pos < c_rli->checkpoint_group; pos++)
{
if (bitmap_is_set(&group_shifted, pos))
bitmap_set_bit(&group_executed, pos - ptr_g->shifted);
}
}
/*
Extracts an updated relay-log name to store in Worker's rli.
*/
if (ptr_g->group_relay_log_name)
{
DBUG_ASSERT(strlen(ptr_g->group_relay_log_name) + 1
<= sizeof(group_relay_log_name));
strmake(group_relay_log_name, ptr_g->group_relay_log_name,
sizeof(group_relay_log_name) - 1);
}
DBUG_ASSERT(ptr_g->checkpoint_seqno <= (c_rli->checkpoint_group - 1));
bitmap_set_bit(&group_executed, ptr_g->checkpoint_seqno);
checkpoint_seqno= ptr_g->checkpoint_seqno;
group_relay_log_pos= ev->future_event_relay_log_pos;
group_master_log_pos= ev->common_header->log_pos;
/*
Directly accessing c_rli->get_group_master_log_name() does not
represent a concurrency issue because the current code places
a synchronization point when master rotates.
*/
strmake(group_master_log_name, c_rli->get_group_master_log_name(),
sizeof(group_master_log_name)-1);
DBUG_PRINT("mts", ("Committing worker-id %lu group master log pos %llu "
"group master log name %s checkpoint sequence number %lu.",
id, group_master_log_pos, group_master_log_name, checkpoint_seqno));
DBUG_EXECUTE_IF("mts_debug_concurrent_access",
{
mts_debug_concurrent_access++;
};
);
DBUG_RETURN(flush_info(force));
}
void Slave_worker::rollback_positions(Slave_job_group* ptr_g)
{
if (!is_transactional())
{
bitmap_clear_bit(&group_executed, ptr_g->checkpoint_seqno);
flush_info(false);
}
}
extern "C" uchar *get_key(const uchar *record, size_t *length,
my_bool not_used MY_ATTRIBUTE((unused)))
{
DBUG_ENTER("get_key");
db_worker_hash_entry *entry=(db_worker_hash_entry *) record;
*length= strlen(entry->db);
DBUG_PRINT("info", ("get_key %s, %d", entry->db, (int) *length));
DBUG_RETURN((uchar*) entry->db);
}
static void free_entry(db_worker_hash_entry *entry)
{
THD *c_thd= current_thd;
DBUG_ENTER("free_entry");
DBUG_PRINT("info", ("free_entry %s, %zu", entry->db, strlen(entry->db)));
DBUG_ASSERT(c_thd->system_thread == SYSTEM_THREAD_SLAVE_SQL);
/*
Although assert is correct valgrind senses entry->worker can be freed.
DBUG_ASSERT(entry->usage == 0 ||
!entry->worker || // last entry owner could have errored out
entry->worker->running_status != Slave_worker::RUNNING);
*/
mts_move_temp_tables_to_thd(c_thd, entry->temporary_tables);
entry->temporary_tables= NULL;
my_free((void *) entry->db);
my_free(entry);
DBUG_VOID_RETURN;
}
bool init_hash_workers(Relay_log_info *rli)
{
DBUG_ENTER("init_hash_workers");
rli->inited_hash_workers=
(my_hash_init(&rli->mapping_db_to_worker, &my_charset_bin,
0, 0, 0, get_key,
(my_hash_free_key) free_entry, 0,
key_memory_db_worker_hash_entry) == 0);
if (rli->inited_hash_workers)
{
mysql_mutex_init(key_mutex_slave_worker_hash, &rli->slave_worker_hash_lock,
MY_MUTEX_INIT_FAST);
mysql_cond_init(key_cond_slave_worker_hash, &rli->slave_worker_hash_cond);
}
DBUG_RETURN (!rli->inited_hash_workers);
}
void destroy_hash_workers(Relay_log_info *rli)
{
DBUG_ENTER("destroy_hash_workers");
if (rli->inited_hash_workers)
{
my_hash_free(&rli->mapping_db_to_worker);
mysql_mutex_destroy(&rli->slave_worker_hash_lock);
mysql_cond_destroy(&rli->slave_worker_hash_cond);
rli->inited_hash_workers= false;
}
DBUG_VOID_RETURN;
}
/**
Relocating temporary table reference into @c entry's table list head.
Sources can be the coordinator's and the Worker's thd->temporary_tables.
@param table TABLE instance pointer
@param thd THD instance pointer of the source of relocation
@param entry db_worker_hash_entry instance pointer
@note thd->temporary_tables can become NULL
@return the pointer to a table following the unlinked
*/
TABLE* mts_move_temp_table_to_entry(TABLE *table, THD *thd,
db_worker_hash_entry *entry)
{
TABLE *ret= table->next;
if (table->prev)
{
table->prev->next= table->next;
if (table->prev->next)
table->next->prev= table->prev;
}
else
{
/* removing the first item from the list */
DBUG_ASSERT(table == thd->temporary_tables);
thd->temporary_tables= table->next;
if (thd->temporary_tables)
table->next->prev= 0;
}
table->next= entry->temporary_tables;
table->prev= 0;
if (table->next)
table->next->prev= table;
entry->temporary_tables= table;
return ret;
}
/**
Relocation of the list of temporary tables to thd->temporary_tables.
@param thd THD instance pointer of the destination
@param temporary_tables
the source temporary_tables list
@note destroying references to the source list, if necessary,
is left to the caller.
@return the post-merge value of thd->temporary_tables.
*/
TABLE* mts_move_temp_tables_to_thd(THD *thd, TABLE *temporary_tables)
{
DBUG_ENTER ("mts_move_temp_tables_to_thd");
TABLE *table= temporary_tables;
if (!table)
DBUG_RETURN(NULL);
// accept only if this is the start of the list.
DBUG_ASSERT(!table->prev);
// walk along the source list and associate the tables with thd
do
{
table->in_use= thd;
} while(table->next && (table= table->next));
// link the former list against the tail of the source list
if (thd->temporary_tables)
thd->temporary_tables->prev= table;
table->next= thd->temporary_tables;
thd->temporary_tables= temporary_tables;
DBUG_RETURN(thd->temporary_tables);
}
/**
Relocating references of temporary tables of a database
of the entry argument from THD into the entry.
@param thd THD pointer of the source temporary_tables list
@param entry a pointer to db_worker_hash_entry record
containing database descriptor and temporary_tables list.
*/
static void move_temp_tables_to_entry(THD* thd, db_worker_hash_entry* entry)
{
for (TABLE *table= thd->temporary_tables; table;)
{
if (strcmp(table->s->db.str, entry->db) == 0)
{
// table pointer is shifted inside the function
table= mts_move_temp_table_to_entry(table, thd, entry);
}
else
{
table= table->next;
}
}
}
/**
The function produces a reference to the struct of a Worker
that has been or will be engaged to process the @c dbname -keyed partition (D).
It checks a local to Coordinator CGAP list first and returns
@c last_assigned_worker when found (todo: assert).
Otherwise, the partition is appended to the current group list:
CGAP .= D
here .= is concatenate operation,
and a possible D's Worker id is searched in Assigned Partition Hash
(APH) that collects tuples (P, W_id, U, mutex, cond).
In case not found,
W_d := W_c unless W_c is NULL.
When W_c is NULL it is assigned to a least occupied as defined by
@c get_least_occupied_worker().
W_d := W_c := W_{least_occupied}
APH .= a new (D, W_d, 1)
In a case APH contains W_d == W_c, (assert U >= 1)
update APH set U++ where APH.P = D
The case APH contains a W_d != W_c != NULL assigned to D-partition represents
the hashing conflict and is handled as the following:
a. marks the record of APH with a flag requesting to signal in the
cond var when `U' the usage counter drops to zero by the other Worker;
b. waits for the other Worker to finish tasks on that partition and
gets the signal;
c. updates the APH record to point to the first Worker (naturally, U := 1),
scheduled the event, and goes back into the parallel mode
@param dbname pointer to c-string containing database name
It can be empty string to indicate specific locking
to faciliate sequential applying.
@param rli pointer to Coordinators relay-log-info instance
@param ptr_entry reference to a pointer to the resulted entry in
the Assigne Partition Hash where
the entry's pointer is stored at return.
@param need_temp_tables
if FALSE migration of temporary tables not needed
@param last_worker caller opts for this Worker, it must be
rli->last_assigned_worker if one is determined.
@note modifies CGAP, APH and unlinks @c dbname -keyd temporary tables
from C's thd->temporary_tables to move them into the entry record.
@return the pointer to a Worker struct
*/
Slave_worker *map_db_to_worker(const char *dbname, Relay_log_info *rli,
db_worker_hash_entry **ptr_entry,
bool need_temp_tables, Slave_worker *last_worker)
{
Slave_worker_array *workers= &rli->workers;
THD *thd= rli->info_thd;
DBUG_ENTER("map_db_to_worker");
DBUG_ASSERT(!rli->last_assigned_worker ||
rli->last_assigned_worker == last_worker);
DBUG_ASSERT(is_mts_db_partitioned(rli));
if (!rli->inited_hash_workers)
DBUG_RETURN(NULL);
db_worker_hash_entry *entry= NULL;
my_hash_value_type hash_value;
size_t dblength= strlen(dbname);
// Search in CGAP
for (db_worker_hash_entry **it= rli->curr_group_assigned_parts.begin();
it != rli->curr_group_assigned_parts.end(); ++it)
{
entry= *it;
if ((uchar) entry->db_len != dblength)
continue;
else
if (strncmp(entry->db, const_cast<char*>(dbname), dblength) == 0)
{
*ptr_entry= entry;
DBUG_RETURN(last_worker);
}
}
DBUG_PRINT("info", ("Searching for %s, %zu", dbname, dblength));
hash_value= my_calc_hash(&rli->mapping_db_to_worker, (uchar*) dbname,
dblength);
mysql_mutex_lock(&rli->slave_worker_hash_lock);
entry= (db_worker_hash_entry *)
my_hash_search_using_hash_value(&rli->mapping_db_to_worker, hash_value,
(uchar*) dbname, dblength);
if (!entry)
{
/*
The database name was not found which means that a worker never
processed events from that database. In such case, we need to
map the database to a worker my inserting an entry into the
hash map.
*/
my_bool ret;
char *db= NULL;
mysql_mutex_unlock(&rli->slave_worker_hash_lock);
DBUG_PRINT("info", ("Inserting %s, %zu", dbname, dblength));
/*
Allocate an entry to be inserted and if the operation fails
an error is returned.
*/
if (!(db= (char *) my_malloc(key_memory_db_worker_hash_entry,
dblength + 1, MYF(0))))
goto err;
if (!(entry= (db_worker_hash_entry *)
my_malloc(key_memory_db_worker_hash_entry,
sizeof(db_worker_hash_entry), MYF(0))))
{
my_free(db);
goto err;
}
my_stpcpy(db, dbname);
entry->db= db;
entry->db_len= strlen(db);
entry->usage= 1;
entry->temporary_tables= NULL;
/*
Unless \exists the last assigned Worker, get a free worker based
on a policy described in the function get_least_occupied_worker().
*/
mysql_mutex_lock(&rli->slave_worker_hash_lock);
entry->worker= (!last_worker) ?
get_least_occupied_worker(rli, workers, NULL) : last_worker;
entry->worker->usage_partition++;
if (rli->mapping_db_to_worker.records > mts_partition_hash_soft_max)
{
/*
A dynamic array to store the mapping_db_to_worker hash elements
that needs to be deleted, since deleting the hash entires while
iterating over it is wrong.
*/
Prealloced_array<db_worker_hash_entry*, HASH_DYNAMIC_INIT>
hash_element(key_memory_db_worker_hash_entry);
/*
remove zero-usage (todo: rare or long ago scheduled) records.
Store the element of the hash in a dynamic array after checking whether
the usage of the hash entry is 0 or not. We later free it from the HASH.
*/
for (uint i= 0; i < rli->mapping_db_to_worker.records; i++)
{
DBUG_ASSERT(!entry->temporary_tables || !entry->temporary_tables->prev);
DBUG_ASSERT(!thd->temporary_tables || !thd->temporary_tables->prev);
db_worker_hash_entry *entry=
(db_worker_hash_entry*) my_hash_element(&rli->mapping_db_to_worker, i);
if (entry->usage == 0)
{
mts_move_temp_tables_to_thd(thd, entry->temporary_tables);
entry->temporary_tables= NULL;
/* Push the element in the dynamic array*/
hash_element.push_back(entry);
}
}
/* Delete the hash element based on the usage */
for (size_t i=0 ; i < hash_element.size(); i++)
{
db_worker_hash_entry *temp_entry= hash_element[i];
my_hash_delete(&rli->mapping_db_to_worker, (uchar*) temp_entry);
}
}
ret= my_hash_insert(&rli->mapping_db_to_worker, (uchar*) entry);
if (ret)
{
my_free(db);
my_free(entry);
entry= NULL;
goto err;
}
DBUG_PRINT("info", ("Inserted %s, %zu", entry->db, strlen(entry->db)));
}
else
{
/* There is a record. Either */
if (entry->usage == 0)
{
entry->worker= (!last_worker) ?
get_least_occupied_worker(rli, workers, NULL) : last_worker;
entry->worker->usage_partition++;
entry->usage++;
}
else if (entry->worker == last_worker || !last_worker)
{
DBUG_ASSERT(entry->worker);
entry->usage++;
}
else
{
// The case APH contains a W_d != W_c != NULL assigned to
// D-partition represents
// the hashing conflict and is handled as the following:
PSI_stage_info old_stage;
DBUG_ASSERT(last_worker != NULL &&
rli->curr_group_assigned_parts.size() > 0);
// future assignenment and marking at the same time
entry->worker= last_worker;
// loop while a user thread is stopping Coordinator gracefully
do
{
thd->ENTER_COND(&rli->slave_worker_hash_cond,
&rli->slave_worker_hash_lock,
&stage_slave_waiting_worker_to_release_partition,
&old_stage);
mysql_cond_wait(&rli->slave_worker_hash_cond, &rli->slave_worker_hash_lock);
} while (entry->usage != 0 && !thd->killed);
mysql_mutex_unlock(&rli->slave_worker_hash_lock);
thd->EXIT_COND(&old_stage);
if (thd->killed)
{
entry= NULL;
goto err;
}
mysql_mutex_lock(&rli->slave_worker_hash_lock);
entry->usage= 1;
entry->worker->usage_partition++;
}
}
/*
relocation belonging to db temporary tables from C to W via entry
*/
if (entry->usage == 1 && need_temp_tables)
{
if (!entry->temporary_tables)
{
if (entry->db_len != 0)
{
move_temp_tables_to_entry(thd, entry);
}
else
{
entry->temporary_tables= thd->temporary_tables;
thd->temporary_tables= NULL;
}
}
#ifndef DBUG_OFF
else
{
// all entries must have been emptied from temps by the caller
for (TABLE *table= thd->temporary_tables; table; table= table->next)
{
DBUG_ASSERT(0 != strcmp(table->s->db.str, entry->db));
}
}
#endif
}
mysql_mutex_unlock(&rli->slave_worker_hash_lock);
DBUG_ASSERT(entry);
err:
if (entry)
{
DBUG_PRINT("info",
("Updating %s with worker %lu", entry->db, entry->worker->id));
rli->curr_group_assigned_parts.push_back(entry);
*ptr_entry= entry;
}
DBUG_RETURN(entry ? entry->worker : NULL);
}
/**
Get the least occupied worker.
@param ws dynarray of pointers to Slave_worker
@return a pointer to chosen Slave_worker instance
*/
Slave_worker *get_least_occupied_worker(Relay_log_info *rli,
Slave_worker_array *ws,
Log_event* ev)
{
return rli->current_mts_submode->get_least_occupied_worker(rli, ws, ev);
}
/**
Deallocation routine to cancel out few effects of
@c map_db_to_worker().
Involved into processing of the group APH tuples are updated.
@c last_group_done_index member is set to the GAQ index of
the current group.
CGEP the Worker partition cache is cleaned up.
@param ev a pointer to Log_event
@param error error code after processing the event by caller.
*/
void Slave_worker::slave_worker_ends_group(Log_event* ev, int error)
{
DBUG_ENTER("Slave_worker::slave_worker_ends_group");
Slave_job_group *ptr_g= NULL;
if (!error)
{
ptr_g= c_rli->gaq->get_job_group(gaq_index);
DBUG_ASSERT(gaq_index == ev->mts_group_idx);
/*
It guarantees that the worker is removed from order commit queue when
its transaction doesn't binlog anything. It will break innodb group commit,
but it should rarely happen.
*/
if (get_commit_order_manager())
get_commit_order_manager()->report_commit(this);
// first ever group must have relay log name
DBUG_ASSERT(last_group_done_index != c_rli->gaq->size ||
ptr_g->group_relay_log_name != NULL);
DBUG_ASSERT(ptr_g->worker_id == id);
if (ev->get_type_code() != binary_log::XID_EVENT)
{
commit_positions(ev, ptr_g, false);
DBUG_EXECUTE_IF("crash_after_commit_and_update_pos",
sql_print_information("Crashing crash_after_commit_and_update_pos.");
flush_info(TRUE);
DBUG_SUICIDE();
);
}
ptr_g->group_master_log_pos= group_master_log_pos;
ptr_g->group_relay_log_pos= group_relay_log_pos;
my_atomic_store32(&ptr_g->done, 1);
last_group_done_index= gaq_index;
last_groups_assigned_index= ptr_g->total_seqno;
reset_gaq_index();
groups_done++;
}
else
{
if (running_status != STOP_ACCEPTED)
{
// tagging as exiting so Coordinator won't be able synchronize with it
mysql_mutex_lock(&jobs_lock);
running_status= ERROR_LEAVING;
mysql_mutex_unlock(&jobs_lock);
/* Fatal error happens, it notifies the following transaction to rollback */
if (get_commit_order_manager())
get_commit_order_manager()->report_rollback(this);
// Killing Coordinator to indicate eventual consistency error
mysql_mutex_lock(&c_rli->info_thd->LOCK_thd_data);
c_rli->info_thd->awake(THD::KILL_QUERY);
mysql_mutex_unlock(&c_rli->info_thd->LOCK_thd_data);
}
}
/*
Cleanup relating to the last executed group regardless of error.
*/
if (current_mts_submode->get_type() == MTS_PARALLEL_TYPE_DB_NAME)
{
for (size_t i= 0; i < curr_group_exec_parts.size(); i++)
{
db_worker_hash_entry *entry= curr_group_exec_parts[i];
mysql_mutex_lock(&c_rli->slave_worker_hash_lock);
DBUG_ASSERT(entry);
entry->usage --;
DBUG_ASSERT(entry->usage >= 0);
if (entry->usage == 0)
{
usage_partition--;
/*
The detached entry's temp table list, possibly updated, remains
with the entry at least until time Coordinator will deallocate it
from the hash, that is either due to stop or extra size of the hash.
*/
DBUG_ASSERT(usage_partition >= 0);
DBUG_ASSERT(this->info_thd->temporary_tables == 0);
DBUG_ASSERT(!entry->temporary_tables ||
!entry->temporary_tables->prev);
if (entry->worker != this) // Coordinator is waiting
{
#ifndef DBUG_OFF
// TODO: open it! DBUG_ASSERT(usage_partition || !entry->worker->jobs.len);
#endif
DBUG_PRINT("info",
("Notifying entry %p release by worker %lu", entry, this->id));
mysql_cond_signal(&c_rli->slave_worker_hash_cond);
}
}
else
DBUG_ASSERT(usage_partition != 0);
mysql_mutex_unlock(&c_rli->slave_worker_hash_lock);
}
curr_group_exec_parts.clear();
curr_group_exec_parts.shrink_to_fit();
if (error)
{
// Awakening Coordinator that could be waiting for entry release
mysql_mutex_lock(&c_rli->slave_worker_hash_lock);
mysql_cond_signal(&c_rli->slave_worker_hash_cond);
mysql_mutex_unlock(&c_rli->slave_worker_hash_lock);
}
}
else // not DB-type scheduler
{
DBUG_ASSERT(current_mts_submode->get_type() ==
MTS_PARALLEL_TYPE_LOGICAL_CLOCK);
/*
Check if there're any waiter. If there're try incrementing lwm and
signal to those who've got sasfied with the waiting condition.
In a "good" "likely" execution branch the waiter set is expected
to be empty. LWM is advanced by Coordinator asynchronously.
Also lwm is advanced by a dependent Worker when it inserts its waiting
request into the waiting list.
*/
Mts_submode_logical_clock* mts_submode=
static_cast<Mts_submode_logical_clock*>(c_rli->current_mts_submode);
int64 min_child_waited_logical_ts=
my_atomic_load64(&mts_submode->min_waited_timestamp);
DBUG_EXECUTE_IF("slave_worker_ends_group_before_signal_lwm",
{
const char act[]= "now WAIT_FOR worker_continue";
DBUG_ASSERT(!debug_sync_set_action(current_thd,
STRING_WITH_LEN(act)));
});
if (unlikely(error))
{
mysql_mutex_lock(&c_rli->mts_gaq_LOCK);
mts_submode->is_error= true;
if (mts_submode->min_waited_timestamp != SEQ_UNINIT)
mysql_cond_signal(&c_rli->logical_clock_cond);
mysql_mutex_unlock(&c_rli->mts_gaq_LOCK);
}
else if (min_child_waited_logical_ts != SEQ_UNINIT)
{
mysql_mutex_lock(&c_rli->mts_gaq_LOCK);
/*
min_child_waited_logical_ts may include an old value, so we need to
check it again after getting the lock.
*/
if (mts_submode->min_waited_timestamp != SEQ_UNINIT)
{
longlong curr_lwm= mts_submode->get_lwm_timestamp(c_rli, true);
if (mts_submode->clock_leq(mts_submode->min_waited_timestamp, curr_lwm))
{
/*
There's a transaction that depends on the current.
*/
mysql_cond_signal(&c_rli->logical_clock_cond);
}
}
mysql_mutex_unlock(&c_rli->mts_gaq_LOCK);
}
#ifndef DBUG_OFF
curr_group_seen_sequence_number= false;
#endif
}
curr_group_seen_gtid= curr_group_seen_begin= false;
DBUG_VOID_RETURN;
}
/**
two index comparision to determine which of the two
is ordered first.
@note The caller makes sure the args are within the valid
range, incl cases the queue is empty or full.
@return TRUE if the first arg identifies a queue entity ordered
after one defined by the 2nd arg,
FALSE otherwise.
*/
template <typename Element_type>
bool circular_buffer_queue<Element_type>::gt(ulong i, ulong k)
{
DBUG_ASSERT(i < size && k < size);
DBUG_ASSERT(avail != entry);
if (i >= entry)
if (k >= entry)
return i > k;
else
return FALSE;
else
if (k >= entry)
return TRUE;
else
return i > k;
}
#ifndef DBUG_OFF
bool Slave_committed_queue::count_done(Relay_log_info* rli)
{
ulong i, k, cnt= 0;
for (i= entry, k= 0; k < len; i= (i + 1) % size, k++)
{
Slave_job_group *ptr_g;
ptr_g= &m_Q[i];
if (ptr_g->worker_id != MTS_WORKER_UNDEF && ptr_g->done)
cnt++;
}
DBUG_ASSERT(cnt <= size);
DBUG_PRINT("mts", ("Checking if it can simulate a crash:"
" mts_checkpoint_group %u counter %lu parallel slaves %lu\n",
opt_mts_checkpoint_group, cnt, rli->slave_parallel_workers));
return (cnt == (rli->slave_parallel_workers * opt_mts_checkpoint_group));
}
#endif
/**
The queue is processed from the head item by item
to purge items representing committed groups.
Progress in GAQ is assessed through comparision of GAQ index value
with Worker's @c last_group_done_index.
Purging breaks at a first discovered gap, that is an item
that the assinged item->w_id'th Worker has not yet completed.
The caller is supposed to be the checkpoint handler.
A copy of the last discarded item containing
the refreshed value of the committed low-water-mark is stored
into @c lwm container member for further caller's processing.
@c last_done is updated with the latest total_seqno for each Worker
that was met during GAQ parse.
@note dyn-allocated members of Slave_job_group such as
group_relay_log_name as freed here.
@return number of discarded items
*/
ulong Slave_committed_queue::move_queue_head(Slave_worker_array *ws)
{
DBUG_ENTER("Slave_committed_queue::move_queue_head");
ulong i, cnt= 0;
for (i= entry; i != avail && !empty(); cnt++, i= (i + 1) % size)
{
Slave_worker *w_i;
Slave_job_group *ptr_g;
char grl_name[FN_REFLEN];
#ifndef DBUG_OFF
if (DBUG_EVALUATE_IF("check_slave_debug_group", 1, 0) &&
cnt == opt_mts_checkpoint_period)
DBUG_RETURN(cnt);
#endif
grl_name[0]= 0;
ptr_g= &m_Q[i];
/*
The current job has not been processed or it was not
even assigned, this means there is a gap.
*/
if (ptr_g->worker_id == MTS_WORKER_UNDEF ||
my_atomic_load32(&ptr_g->done) == 0)
break; /* gap at i'th */
/* Worker-id domain guard */
compile_time_assert(MTS_WORKER_UNDEF > MTS_MAX_WORKERS);
w_i= ws->at(ptr_g->worker_id);
/*
Memorizes the latest valid group_relay_log_name.
*/
if (ptr_g->group_relay_log_name)
{
strcpy(grl_name, ptr_g->group_relay_log_name);
my_free(ptr_g->group_relay_log_name);
/*
It is important to mark the field as freed.
*/
ptr_g->group_relay_log_name= NULL;
}
/*
Removes the job from the (G)lobal (A)ssigned (Q)ueue.
*/
Slave_job_group g= Slave_job_group();
#ifndef DBUG_OFF
ulong ind=
#endif
de_queue(&g);
/*
Stores the memorized name into the result struct. Note that we
take care of the pointer first and then copy the other elements
by assigning the structures.
*/
if (grl_name[0] != 0)
{
strcpy(lwm.group_relay_log_name, grl_name);
}
g.group_relay_log_name= lwm.group_relay_log_name;
lwm= g;
DBUG_ASSERT(ind == i);
DBUG_ASSERT(!ptr_g->group_relay_log_name);
DBUG_ASSERT(ptr_g->total_seqno == lwm.total_seqno);
#ifndef DBUG_OFF
{
ulonglong l= last_done[w_i->id];
/*
There must be some progress otherwise we should have
exit the loop earlier.
*/
DBUG_ASSERT(l < ptr_g->total_seqno);
}
#endif
/*
This is used to calculate the last time each worker has
processed events.
*/
last_done[w_i->id]= ptr_g->total_seqno;
}
DBUG_ASSERT(cnt <= size);
DBUG_RETURN(cnt);
}
/**
Finds low-water mark of committed jobs in GAQ.
That is an index below which all jobs are marked as done.
Notice the first available index is returned when the queue
does not have any incomplete jobs. That includes cases of
the empty and the full of complete jobs queue.
A mutex protecting from concurrent LWM change by
move_queue_head() (by Coordinator) should be taken by the caller.
@param arg_g [out] a double pointer to Slave job descriptor item
last marked with done-as-true boolean.
@param start_index a GAQ index to start/resume searching.
Caller is to make sure the index points into
assigned (occupied) range of circular buffer of GAQ.
@return GAQ index of the last consecutive done job, or the GAQ
size when none is found.
*/
ulong Slave_committed_queue::find_lwm(Slave_job_group** arg_g,
ulong start_index)
{
Slave_job_group *ptr_g= NULL;
ulong i, k, cnt;
DBUG_ASSERT(start_index <= size);
if (empty())
return size;
/*
Loop continuation condition relies on
(TODO: assert it)
the start_index being in the running range:
start_index \in [entry, avail - 1].
It satisfies any queue size including 1.
It does not satisfy the empty queue case which is bailed out earlier above.
*/
for (i= start_index, cnt= 0; cnt < len - (start_index + size - entry) % size;
i= (i + 1) % size, cnt++)
{
ptr_g= &m_Q[i];
if (my_atomic_load32(&ptr_g->done) == 0)
{
if (cnt == 0)
return size; // the first node of the queue is not done
break;
}
}
ptr_g= &m_Q[k= (i + size - 1) % size];
*arg_g= ptr_g;
return k;
}
/**
Method should be executed at slave system stop to
cleanup dynamically allocated items that remained as unprocessed
by Coordinator and Workers in their regular execution course.
*/
void Slave_committed_queue::free_dynamic_items()
{
ulong i, k;
for (i= entry, k= 0; k < len; i= (i + 1) % size, k++)
{
Slave_job_group *ptr_g= &m_Q[i];
if (ptr_g->group_relay_log_name)
{
my_free(ptr_g->group_relay_log_name);
}
if (ptr_g->checkpoint_log_name)
{
my_free(ptr_g->checkpoint_log_name);
}
if (ptr_g->checkpoint_relay_log_name)
{
my_free(ptr_g->checkpoint_relay_log_name);
}
if (ptr_g->group_master_log_name)
{
my_free(ptr_g->group_master_log_name);
}
}
DBUG_ASSERT((avail == size /* full */ || entry == size /* empty */) ||
i == avail /* all occupied are processed */);
}
void Slave_worker::do_report(loglevel level, int err_code, const char *msg,
va_list args) const
{
char buff_coord[MAX_SLAVE_ERRMSG];
char buff_gtid[Gtid::MAX_TEXT_LENGTH + 1];
const char* log_name= const_cast<Slave_worker*>(this)->get_master_log_name();
ulonglong log_pos= const_cast<Slave_worker*>(this)->get_master_log_pos();
bool is_group_replication_applier_channel=
channel_map.is_group_replication_channel_name(c_rli->get_channel(), true);
const Gtid_specification *gtid_next= &info_thd->variables.gtid_next;
THD *thd= info_thd;
gtid_next->to_string(global_sid_map, buff_gtid, true);
if (level == ERROR_LEVEL && (!has_temporary_error(thd, err_code) ||
thd->get_transaction()->cannot_safely_rollback(Transaction_ctx::SESSION)))
{
char coordinator_errmsg[MAX_SLAVE_ERRMSG];
if (is_group_replication_applier_channel)
{
my_snprintf(coordinator_errmsg, MAX_SLAVE_ERRMSG,
"Coordinator stopped because there were error(s) in the "
"worker(s). "
"The most recent failure being: Worker %u failed executing "
"transaction '%s'. See error log and/or "
"performance_schema.replication_applier_status_by_worker "
"table for "
"more details about this failure or others, if any.",
internal_id, buff_gtid);
}
else
{
my_snprintf(coordinator_errmsg, MAX_SLAVE_ERRMSG,
"Coordinator stopped because there were error(s) in the "
"worker(s). "
"The most recent failure being: Worker %u failed executing "
"transaction '%s' at master log %s, end_log_pos %llu. "
"See error log and/or "
"performance_schema.replication_applier_status_by_worker "
"table for "
"more details about this failure or others, if any.",
internal_id, buff_gtid, log_name, log_pos);
}
/*
We want to update the errors in coordinator as well as worker.
The fill_coord_err_buf() function update the error number, message and
timestamp fields. This function is different from va_report() as va_report()
also logs the error message in the log apart from updating the error fields.
So, the worker does the job of reporting the error in the log. We just make
coordinator aware of the error.
*/
c_rli->fill_coord_err_buf(level, err_code, coordinator_errmsg);
}
if (is_group_replication_applier_channel)
{
my_snprintf(buff_coord, sizeof(buff_coord),
"Worker %u failed executing transaction '%s'",
internal_id, buff_gtid);
}
else
{
my_snprintf(buff_coord, sizeof(buff_coord),
"Worker %u failed executing transaction '%s' at "
"master log %s, end_log_pos %llu",
internal_id, buff_gtid, log_name, log_pos);
}
/*
Error reporting by the worker. The worker updates its error fields as well
as reports the error in the log.
*/
this->va_report(level, err_code, buff_coord, msg, args);
}
#ifndef DBUG_OFF
static bool may_have_timestamp(Log_event *ev)
{
bool res= false;
switch (ev->get_type_code())
{
case binary_log::QUERY_EVENT:
res= true;
break;
case binary_log::GTID_LOG_EVENT:
res= true;
break;
default:
break;
}
return res;
}
static int64 get_last_committed(Log_event *ev)
{
int64 res= SEQ_UNINIT;
switch (ev->get_type_code())
{
case binary_log::GTID_LOG_EVENT:
res= static_cast<Gtid_log_event*>(ev)->last_committed;
break;
default:
break;
}
return res;
}
static int64 get_sequence_number(Log_event *ev)
{
int64 res= SEQ_UNINIT;
switch (ev->get_type_code())
{
case binary_log::GTID_LOG_EVENT:
res= static_cast<Gtid_log_event*>(ev)->sequence_number;
break;
default:
break;
}
return res;
}
#endif
/**
MTS worker main routine.
The worker thread loops in waiting for an event, executing it and
fixing statistics counters.
@param worker a pointer to the assigned Worker struct
@param rli a pointer to Relay_log_info of Coordinator
to update statistics.
@return 0 success
-1 got killed or an error happened during appying
*/
int Slave_worker::slave_worker_exec_event(Log_event *ev)
{
Relay_log_info *rli= c_rli;
THD *thd= info_thd;
int ret= 0;
DBUG_ENTER("slave_worker_exec_event");
thd->server_id = ev->server_id;
thd->set_time();
thd->lex->set_current_select(0);
if (!ev->common_header->when.tv_sec)
ev->common_header->when.tv_sec= static_cast<long>(my_time(0));
ev->thd= thd; // todo: assert because up to this point, ev->thd == 0
ev->worker= this;
#ifndef DBUG_OFF
if (!is_mts_db_partitioned(rli) && may_have_timestamp(ev) &&
!curr_group_seen_sequence_number)
{
curr_group_seen_sequence_number= true;
longlong lwm_estimate= static_cast<Mts_submode_logical_clock*>
(rli->current_mts_submode)->estimate_lwm_timestamp();
int64 last_committed= get_last_committed(ev);
int64 sequence_number= get_sequence_number(ev);
/*
The commit timestamp waiting condition:
lwm_estimate < last_committed <=> last_committed \not <= lwm_estimate
must have been satisfied by Coordinator.
The first scheduled transaction does not have to wait for anybody.
*/
DBUG_ASSERT(rli->gaq->entry == ev->mts_group_idx ||
Mts_submode_logical_clock::clock_leq(last_committed,
lwm_estimate));
DBUG_ASSERT(lwm_estimate != SEQ_UNINIT || rli->gaq->entry == ev->mts_group_idx);
/*
The current transaction's timestamp can't be less that lwm.
*/
DBUG_ASSERT(sequence_number == SEQ_UNINIT ||
!Mts_submode_logical_clock::
clock_leq(sequence_number,
static_cast<Mts_submode_logical_clock*>
(rli->current_mts_submode)->
estimate_lwm_timestamp()));
}
#endif
// Address partioning only in database mode
if (!is_gtid_event(ev) && is_mts_db_partitioned(rli))
{
if (ev->contains_partition_info(end_group_sets_max_dbs))
{
uint num_dbs= ev->mts_number_dbs();
if (num_dbs == OVER_MAX_DBS_IN_EVENT_MTS)
num_dbs= 1;
DBUG_ASSERT(num_dbs > 0);
for (uint k= 0; k < num_dbs; k++)
{
bool found= false;
for (size_t i= 0; i < curr_group_exec_parts.size() && !found; i++)
{
found= curr_group_exec_parts[i] ==
ev->mts_assigned_partitions[k];
}
if (!found)
{
/*
notice, can't assert
DBUG_ASSERT(ev->mts_assigned_partitions[k]->worker == worker);
since entry could be marked as wanted by other worker.
*/
curr_group_exec_parts.push_back(ev->mts_assigned_partitions[k]);
}
}
end_group_sets_max_dbs= false;
}
}
set_future_event_relay_log_pos(ev->future_event_relay_log_pos);
set_master_log_pos(static_cast<ulong>(ev->common_header->log_pos));
set_gaq_index(ev->mts_group_idx);
ret= ev->do_apply_event_worker(this);
DBUG_RETURN(ret);
}
/**
Sleep for a given amount of seconds or until killed.
@param seconds The number of seconds to sleep.
@retval True if the thread has been killed, false otherwise.
*/
bool Slave_worker::worker_sleep(ulong seconds)
{
bool ret= false;
struct timespec abstime;
mysql_mutex_t *lock= &jobs_lock;
mysql_cond_t *cond= &jobs_cond;
/* Absolute system time at which the sleep time expires. */
set_timespec(&abstime, seconds);
mysql_mutex_lock(lock);
info_thd->ENTER_COND(cond, lock, NULL, NULL);
while (!(ret= info_thd->killed || running_status != RUNNING))
{
int error= mysql_cond_timedwait(cond, lock, &abstime);
if (error == ETIMEDOUT || error == ETIME)
break;
}
mysql_mutex_unlock(lock);
info_thd->EXIT_COND(NULL);
return ret;
}
/**
It is called after an error happens. It checks if that is an temporary
error and if the situation is allow to retry the transaction. Then it will
retry the transaction if it is allowed. Retry policy and logic is similar to
single-threaded slave.
@param[in] start_relay_number The extension number of the relay log which
includes the first event of the transaction.
@param[in] start_relay_pos The offset of the transaction's first event.
@param[in] end_relay_number The extension number of the relay log which
includes the last event it should retry.
@param[in] end_relay_pos The offset of the last event it should retry.
@return false if succeeds, otherwise returns true.
*/
bool Slave_worker::retry_transaction(uint start_relay_number,
my_off_t start_relay_pos,
uint end_relay_number,
my_off_t end_relay_pos)
{
THD *thd= info_thd;
bool silent= false;
DBUG_ENTER("Slave_worker::retry_transaction");
if (slave_trans_retries == 0)
DBUG_RETURN(true);
do
{
/* Simulate a lock deadlock error */
uint error= 0;
if (found_order_commit_deadlock())
{
/*
This transaction was allowed to be executed in parallel with other that
happened earlier according to binary log order. It was asked to be
rolled back by the other transaction as it was holding a lock that is
needed by the other transaction to progress, according to binary log
order this configure a deadlock.
At this point, this transaction *should* have no non-temporary errors.
Having a non-temporary error may be a sign of:
a) Slave has diverged from the master;
b) There is an issue in the logical clock allowing a transaction to be
applied in parallel with its dependencies (the two transactions are
trying to change the same record in parallel).
For (a), a retry of this transaction will produce the same error. For
(b), this transaction might succeed upon retry, allowing the slave to
progress without manual intervention, but it is a sign of problems in LC
generation at the master.
So, we will make the worker to retry this transaction only if there is
no error or the error is a temporary error.
*/
Diagnostics_area *da= thd->get_stmt_da();
if (!thd->get_stmt_da()->is_error() ||
has_temporary_error(thd,
da->is_error() ? da->mysql_errno() : error,
&silent))
{
error= ER_LOCK_DEADLOCK;
}
#ifndef DBUG_OFF
else
{
/*
The non-debug binary will not retry this transactions, stopping the
SQL thread because of the non-temporary error. But, as this situation
is not supposed to happen as described in the comment above, we will
fail an assert to ease the issue investigation when it happens.
*/
if (DBUG_EVALUATE_IF("rpl_fake_cod_deadlock", 0, 1))
DBUG_ASSERT(false);
}
#endif
}
if (!has_temporary_error(thd, error, &silent) ||
thd->get_transaction()->cannot_safely_rollback(Transaction_ctx::SESSION))
DBUG_RETURN(true);
if (trans_retries >= slave_trans_retries)
{
thd->is_fatal_error= 1;
c_rli->report(ERROR_LEVEL, thd->get_stmt_da()->mysql_errno(),
"worker thread retried transaction %lu time(s) "
"in vain, giving up. Consider raising the value of "
"the slave_transaction_retries variable.", trans_retries);
DBUG_RETURN(true);
}
if (!silent)
trans_retries++;
mysql_mutex_lock(&c_rli->data_lock);
c_rli->retried_trans++;
mysql_mutex_unlock(&c_rli->data_lock);
cleanup_context(thd, 1);
reset_order_commit_deadlock();
worker_sleep(min<ulong>(trans_retries, MAX_SLAVE_RETRY_PAUSE));
} while (read_and_apply_events(start_relay_number, start_relay_pos,
end_relay_number, end_relay_pos));
DBUG_RETURN(false);
}
/**
Read events from relay logs and apply them.
@param[in] start_relay_number The extension number of the relay log which
includes the first event of the transaction.
@param[in] start_relay_pos The offset of the transaction's first event.
@param[in] end_relay_number The extension number of the relay log which
includes the last event it should retry.
@param[in] end_relay_pos The offset of the last event it should retry.
@return false if succeeds, otherwise returns true.
*/
bool Slave_worker::read_and_apply_events(uint start_relay_number,
my_off_t start_relay_pos,
uint end_relay_number,
my_off_t end_relay_pos)
{
DBUG_ENTER("Slave_worker::read_and_apply_events");
Relay_log_info *rli= c_rli;
IO_CACHE relay_io;
char file_name[FN_REFLEN+1];
uint file_number= start_relay_number;
bool error= true;
bool arrive_end= false;
relay_log_number_to_name(start_relay_number, file_name);
memset(&relay_io, 0, sizeof(IO_CACHE));
while (!arrive_end)
{
Log_event *ev= NULL;
if (!my_b_inited(&relay_io))
{
const char *errmsg;
DBUG_PRINT("info", ("Open relay log %s", file_name));
if (open_binlog_file(&relay_io, file_name, &errmsg) == -1)
{
sql_print_error("Failed to open relay log %s, error: %s", file_name,
errmsg);
goto end;
}
my_b_seek(&relay_io, start_relay_pos);
}
/* If it is the last event, then set arrive_end as true */
arrive_end= (my_b_tell(&relay_io) == end_relay_pos &&
file_number == end_relay_number);
ev= Log_event::read_log_event(&relay_io, NULL,
rli->get_rli_description_event(),
opt_slave_sql_verify_checksum);
if (ev != NULL)
{
/* It is a event belongs to the transaction */
if (!ev->is_mts_sequential_exec(rli->current_mts_submode->get_type() ==
MTS_PARALLEL_TYPE_DB_NAME))
{
int ret= 0;
ev->future_event_relay_log_pos= my_b_tell(&relay_io);
ev->mts_group_idx= gaq_index;
if (is_mts_db_partitioned(rli) && ev->contains_partition_info(true))
assign_partition_db(ev);
ret= slave_worker_exec_event(ev);
if (ev->worker != NULL)
{
delete ev;
ev= NULL;
}
if (ret != 0)
goto end;
}
else
{
/*
It is a Rotate_log_event, Format_description_log_event event or other
type event doesn't belong to the transaction.
*/
delete ev;
ev= NULL;
}
}
else
{
/*
IO error happens if relay_io.error != 0, otherwise it arrives the
end of the relay log
*/
if (relay_io.error != 0)
{
sql_print_error("Error when worker read relay log events,"
"relay log name %s, position %llu",
rli->get_event_relay_log_name(), my_b_tell(&relay_io));
goto end;
}
if (rli->relay_log.find_next_relay_log(file_name))
{
sql_print_error("Failed to find next relay log when retrying the "
"transaction, current relay log is %s", file_name);
goto end;
}
file_number= relay_log_name_to_number(file_name);
end_io_cache(&relay_io);
mysql_file_close(relay_io.file, MYF(0));
start_relay_pos= BIN_LOG_HEADER_SIZE;
}
}
error= false;
end:
if (my_b_inited(&relay_io))
{
end_io_cache(&relay_io);
mysql_file_close(relay_io.file, MYF(0));
}
DBUG_RETURN(error);
}
/*
Find database entry from map_db_to_worker hash table.
*/
static db_worker_hash_entry *find_entry_from_db_map(const char *dbname,
Relay_log_info *rli)
{
db_worker_hash_entry *entry= NULL;
my_hash_value_type hash_value;
uchar dblength= (uint) strlen(dbname);
hash_value= my_calc_hash(&rli->mapping_db_to_worker, (const uchar*) dbname,
dblength);
mysql_mutex_lock(&rli->slave_worker_hash_lock);
entry= (db_worker_hash_entry *)
my_hash_search_using_hash_value(&rli->mapping_db_to_worker, hash_value,
(uchar*) dbname, dblength);
mysql_mutex_unlock(&rli->slave_worker_hash_lock);
return entry;
}
/*
Initialize Log_event::mts_assigned_partitions array. It is for transaction
retry and is only called when retrying a transaction by workers.
*/
void Slave_worker::assign_partition_db(Log_event *ev)
{
Mts_db_names mts_dbs;
int i;
ev->get_mts_dbs(&mts_dbs);
if (mts_dbs.num == OVER_MAX_DBS_IN_EVENT_MTS)
ev->mts_assigned_partitions[0]= find_entry_from_db_map("", c_rli);
else
for (i= 0; i < mts_dbs.num; i++)
ev->mts_assigned_partitions[i]= find_entry_from_db_map(mts_dbs.name[i],
c_rli);
}
// returns the next available! (TODO: incompatible to circurla_buff method!!!)
static int en_queue(Slave_jobs_queue *jobs, Slave_job_item *item)
{
if (jobs->avail == jobs->size)
{
DBUG_ASSERT(jobs->avail == jobs->m_Q.size());
return -1;
}
// store
jobs->m_Q[jobs->avail]= *item;
// pre-boundary cond
if (jobs->entry == jobs->size)
jobs->entry= jobs->avail;
jobs->avail= (jobs->avail + 1) % jobs->size;
jobs->len++;
// post-boundary cond
if (jobs->avail == jobs->entry)
jobs->avail= jobs->size;
DBUG_ASSERT(jobs->avail == jobs->entry ||
jobs->len == (jobs->avail >= jobs->entry) ?
(jobs->avail - jobs->entry) : (jobs->size + jobs->avail - jobs->entry));
return jobs->avail;
}
/**
return the value of @c data member of the head of the queue.
*/
void * head_queue(Slave_jobs_queue *jobs, Slave_job_item *ret)
{
if (jobs->entry == jobs->size)
{
DBUG_ASSERT(jobs->len == 0);
ret->data= NULL; // todo: move to caller
return NULL;
}
*ret= jobs->m_Q[jobs->entry];
DBUG_ASSERT(ret->data); // todo: move to caller
return ret;
}
/**
return a job item through a struct which point is supplied via argument.
*/
Slave_job_item * de_queue(Slave_jobs_queue *jobs, Slave_job_item *ret)
{
if (jobs->entry == jobs->size)
{
DBUG_ASSERT(jobs->len == 0);
return NULL;
}
*ret= jobs->m_Q[jobs->entry];
jobs->len--;
// pre boundary cond
if (jobs->avail == jobs->size)
jobs->avail= jobs->entry;
jobs->entry= (jobs->entry + 1) % jobs->size;
// post boundary cond
if (jobs->avail == jobs->entry)
jobs->entry= jobs->size;
DBUG_ASSERT(jobs->entry == jobs->size ||
(jobs->len == (jobs->avail >= jobs->entry) ?
(jobs->avail - jobs->entry) :
(jobs->size + jobs->avail - jobs->entry)));
return ret;
}
/**
Coordinator enqueues a job item into a Worker private queue.
@param job_item a pointer to struct carrying a reference to an event
@param worker a pointer to the assigned Worker struct
@param rli a pointer to Relay_log_info of Coordinator
@return false Success.
true Thread killed or worker stopped while waiting for
successful enqueue.
*/
bool append_item_to_jobs(slave_job_item *job_item,
Slave_worker *worker, Relay_log_info *rli)
{
THD *thd= rli->info_thd;
int ret= -1;
size_t ev_size= job_item->data->common_header->data_written;
ulonglong new_pend_size;
PSI_stage_info old_stage;
DBUG_ASSERT(thd == current_thd);
mysql_mutex_lock(&rli->pending_jobs_lock);
new_pend_size= rli->mts_pending_jobs_size + ev_size;
bool big_event= (ev_size > rli->mts_pending_jobs_size_max);
/*
C waits basing on *data* sizes in the queues.
If it is a big event (event size is greater than
slave_pending_jobs_size_max but less than slave_max_allowed_packet),
it will wait for all the jobs in the workers's queue to be
completed. If it is normal event (event size is less than
slave_pending_jobs_size_max), then it will wait for
enough empty memory to keep the event in one of the workers's
queue.
NOTE: Receiver thread (I/O thread) is taking care of restricting
the event size to slave_max_allowed_packet. If an event from
the master is bigger than this value, IO thread will be stopped
with error ER_NET_PACKET_TOO_LARGE.
*/
while ( (!big_event && new_pend_size > rli->mts_pending_jobs_size_max)
|| (big_event && rli->mts_pending_jobs_size != 0 ))
{
rli->mts_wq_oversize= TRUE;
rli->wq_size_waits_cnt++; // waiting due to the total size
thd->ENTER_COND(&rli->pending_jobs_cond, &rli->pending_jobs_lock,
&stage_slave_waiting_worker_to_free_events, &old_stage);
mysql_cond_wait(&rli->pending_jobs_cond, &rli->pending_jobs_lock);
mysql_mutex_unlock(&rli->pending_jobs_lock);
thd->EXIT_COND(&old_stage);
if (thd->killed)
return true;
if (rli->wq_size_waits_cnt % 10 == 1)
sql_print_information("Multi-threaded slave: Coordinator has waited "
"%lu times hitting slave_pending_jobs_size_max; "
"current event size = %zu.",
rli->wq_size_waits_cnt, ev_size);
mysql_mutex_lock(&rli->pending_jobs_lock);
new_pend_size= rli->mts_pending_jobs_size + ev_size;
}
rli->pending_jobs++;
rli->mts_pending_jobs_size= new_pend_size;
rli->mts_events_assigned++;
mysql_mutex_unlock(&rli->pending_jobs_lock);
/*
Sleep unless there is an underrunning Worker and the current Worker
queue is empty or filled lightly (not more than underrun level).
*/
if (rli->mts_wq_underrun_w_id == MTS_WORKER_UNDEF &&
worker->jobs.len > worker->underrun_level)
{
/*
todo: experiment with weight to get a good approximation formula.
Max possible nap time is choosen 1 ms.
The bigger the excessive overrun counter the longer the nap.
*/
ulong nap_weight= rli->mts_wq_excess_cnt + 1;
/*
Nap time is a product of a weight factor and the basic nap unit.
The weight factor is proportional to the worker queues overrun excess
counter. For example when there were only one overruning Worker
the max nap_weight as 0.1 * worker->jobs.size would be
about 1600 so the max nap time is approx 0.008 secs.
Such value is not reachable because of min().
Notice, granularity of sleep depends on the resolution of the software
clock, High-Resolution Timer (HRT) configuration. Without HRT
the precision of wake-up through @c select() may be greater or
equal 1 ms. So don't expect the nap last a prescribed fraction of 1 ms
in such case.
*/
my_sleep(min<ulong>(1000, nap_weight * rli->mts_coordinator_basic_nap));
rli->mts_wq_no_underrun_cnt++;
}
mysql_mutex_lock(&worker->jobs_lock);
// possible WQ overfill
while (worker->running_status == Slave_worker::RUNNING && !thd->killed &&
(ret= en_queue(&worker->jobs, job_item)) == -1)
{
thd->ENTER_COND(&worker->jobs_cond, &worker->jobs_lock,
&stage_slave_waiting_worker_queue, &old_stage);
worker->jobs.overfill= TRUE;
worker->jobs.waited_overfill++;
rli->mts_wq_overfill_cnt++;
mysql_cond_wait(&worker->jobs_cond, &worker->jobs_lock);
mysql_mutex_unlock(&worker->jobs_lock);
thd->EXIT_COND(&old_stage);
mysql_mutex_lock(&worker->jobs_lock);
}
if (ret != -1)
{
worker->curr_jobs++;
if (worker->jobs.len == 1)
mysql_cond_signal(&worker->jobs_cond);
mysql_mutex_unlock(&worker->jobs_lock);
}
else
{
mysql_mutex_unlock(&worker->jobs_lock);
mysql_mutex_lock(&rli->pending_jobs_lock);
rli->pending_jobs--; // roll back of the prev incr
rli->mts_pending_jobs_size -= ev_size;
mysql_mutex_unlock(&rli->pending_jobs_lock);
}
return (-1 != ret ? false : true);
}
/**
Remove a job item from the given workers job queue. It also updates related
status.
param[in] job_item The job item will be removed
param[in] worker The worker which job_item belongs to.
param[in] rli slave's relay log info object.
*/
static void remove_item_from_jobs(slave_job_item *job_item,
Slave_worker *worker, Relay_log_info *rli)
{
Log_event *ev= job_item->data;
mysql_mutex_lock(&worker->jobs_lock);
de_queue(&worker->jobs, job_item);
/* possible overfill */
if (worker->jobs.len == worker->jobs.size - 1 &&
worker->jobs.overfill == TRUE)
{
worker->jobs.overfill= false;
// todo: worker->hungry_cnt++;
mysql_cond_signal(&worker->jobs_cond);
}
mysql_mutex_unlock(&worker->jobs_lock);
/* statistics */
/* todo: convert to rwlock/atomic write */
mysql_mutex_lock(&rli->pending_jobs_lock);
rli->pending_jobs--;
rli->mts_pending_jobs_size-= ev->common_header->data_written;
DBUG_ASSERT(rli->mts_pending_jobs_size < rli->mts_pending_jobs_size_max);
/*
The positive branch is underrun: number of pending assignments
is less than underrun level.
Zero of jobs.len has to reset underrun w_id as the worker may get
the next piece of assignement in a long time.
*/
if (worker->underrun_level > worker->jobs.len && worker->jobs.len != 0)
{
rli->mts_wq_underrun_w_id= worker->id;
} else if (rli->mts_wq_underrun_w_id == worker->id)
{
// reset only own marking
rli->mts_wq_underrun_w_id= MTS_WORKER_UNDEF;
}
/*
Overrun handling.
Incrementing the Worker private and the total excess counter corresponding
to number of events filled above the overrun_level.
The increment amount to the total counter is a difference between
the current and the previous private excess (worker->wq_overrun_cnt).
When the current queue length drops below overrun_level the global
counter is decremented, the local is reset.
*/
if (worker->overrun_level < worker->jobs.len)
{
ulong last_overrun= worker->wq_overrun_cnt;
ulong excess_delta;
/* current overrun */
worker->wq_overrun_cnt= worker->jobs.len - worker->overrun_level;
excess_delta= worker->wq_overrun_cnt - last_overrun;
worker->excess_cnt+= excess_delta;
rli->mts_wq_excess_cnt+= excess_delta;
rli->mts_wq_overrun_cnt++; // statistics
// guarding correctness of incrementing in case of the only one Worker
DBUG_ASSERT(rli->workers.size() != 1 ||
rli->mts_wq_excess_cnt == worker->wq_overrun_cnt);
}
else if (worker->excess_cnt > 0)
{
// When level drops below the total excess is decremented by the
// value of the worker's contribution to the total excess.
rli->mts_wq_excess_cnt-= worker->excess_cnt;
worker->excess_cnt= 0;
worker->wq_overrun_cnt= 0; // and the local is reset
DBUG_ASSERT(rli->mts_wq_excess_cnt >= 0);
DBUG_ASSERT(rli->mts_wq_excess_cnt == 0 || rli->workers.size() > 1);
}
/* coordinator can be waiting */
if (rli->mts_pending_jobs_size < rli->mts_pending_jobs_size_max &&
rli->mts_wq_oversize) // TODO: unit/general test wq_oversize
{
rli->mts_wq_oversize= FALSE;
mysql_cond_signal(&rli->pending_jobs_cond);
}
mysql_mutex_unlock(&rli->pending_jobs_lock);
worker->events_done++;
}
/**
Worker's routine to wait for a new assignement through
@c append_item_to_jobs()
@param worker a pointer to the waiting Worker struct
@param job_item a pointer to struct carrying a reference to an event
@return NULL failure or
a-pointer to an item.
*/
struct slave_job_item* pop_jobs_item(Slave_worker *worker,
Slave_job_item *job_item)
{
THD *thd= worker->info_thd;
mysql_mutex_lock(&worker->jobs_lock);
job_item->data= NULL;
while (!job_item->data && !thd->killed &&
(worker->running_status == Slave_worker::RUNNING ||
worker->running_status == Slave_worker::STOP))
{
PSI_stage_info old_stage;
if (set_max_updated_index_on_stop(worker, job_item))
break;
if (job_item->data == NULL)
{
worker->wq_empty_waits++;
thd->ENTER_COND(&worker->jobs_cond, &worker->jobs_lock,
&stage_slave_waiting_event_from_coordinator,
&old_stage);
mysql_cond_wait(&worker->jobs_cond, &worker->jobs_lock);
mysql_mutex_unlock(&worker->jobs_lock);
thd->EXIT_COND(&old_stage);
mysql_mutex_lock(&worker->jobs_lock);
}
}
if (job_item->data)
worker->curr_jobs--;
mysql_mutex_unlock(&worker->jobs_lock);
thd_proc_info(worker->info_thd, "Executing event");
return job_item;
}
/**
Report a not yet reported error to the coordinator if necessary.
All issues detected when applying binary log events are reported using
rli->report(), but when an issue is not reported by the log event being
applied, there is a workaround at handle_slave_sql() to report the issue
also using rli->report() for the STS applier (or the MTS coordinator).
This function implements the workaround for a MTS worker.
@param worker the worker to be evaluated.
*/
void report_error_to_coordinator(Slave_worker *worker)
{
THD *thd= worker->info_thd;
/*
It is possible that the worker had failed to apply the event but
did not reported about the failure using rli->report(). An example
of such cases are failures caused by setting GTID_NEXT variable with
an unsupported GTID mode (GTID_SET when GTID_MODE = OFF, anonymous
GTID when GTID_MODE = ON).
*/
if (thd->is_error())
{
char const *const errmsg= thd->get_stmt_da()->message_text();
DBUG_PRINT("info",
("thd->get_stmt_da()->get_mysql_errno()=%d; "
"worker->last_error.number=%d",
thd->get_stmt_da()->mysql_errno(),
worker->last_error().number));
if (worker->last_error().number == 0 &&
/*
When another worker that should commit before the current worker
being evaluated has failed and the commit order should be preserved
the current worker was asked to roll back and would stop with the
ER_SLAVE_WORKER_STOPPED_PREVIOUS_THD_ERROR not yet reported to the
coordinator. Reporting this error to the coordinator would be a
mistake and would mask the real issue that lead to the MTS stop as
the coordinator reports only the last error reported to it as the
cause of the MTS failure.
So, we should skip reporting the error if it was reported because
the current transaction had to be rolled back by a failure in a
previous transaction in the commit order while the current
transaction was waiting to be committed.
*/
thd->get_stmt_da()->mysql_errno() !=
ER_SLAVE_WORKER_STOPPED_PREVIOUS_THD_ERROR)
{
/*
This function is reporting an error which was not reported
while executing exec_relay_log_event().
*/
worker->report(ERROR_LEVEL, thd->get_stmt_da()->mysql_errno(),
"%s", errmsg);
}
}
}
/**
apply one job group.
@note the function maintains worker's CGEP and modifies APH, updates
the current group item in GAQ via @c slave_worker_ends_group().
param[in] worker the worker which calls it.
param[in] rli slave's relay log info object.
return returns 0 if the group of jobs are applied successfully, otherwise
returns an error code.
*/
int slave_worker_exec_job_group(Slave_worker *worker, Relay_log_info *rli)
{
struct slave_job_item item= {NULL, 0, 0};
struct slave_job_item *job_item= &item;
THD *thd= worker->info_thd;
bool seen_gtid= false;
bool seen_begin= false;
int error= 0;
Log_event *ev= NULL;
uint start_relay_number;
my_off_t start_relay_pos;
DBUG_ENTER("slave_worker_exec_job_group");
if (unlikely(worker->trans_retries > 0))
worker->trans_retries= 0;
job_item= pop_jobs_item(worker, job_item);
start_relay_number= job_item->relay_number;
start_relay_pos= job_item->relay_pos;
while (1)
{
Slave_job_group *ptr_g;
if (unlikely(thd->killed || worker->running_status == Slave_worker::STOP_ACCEPTED))
{
DBUG_ASSERT(worker->running_status != Slave_worker::ERROR_LEAVING);
// de-queueing and decrement counters is in the caller's exit branch
error= -1;
goto err;
}
ev= job_item->data;
DBUG_ASSERT(ev != NULL);
DBUG_PRINT("info", ("W_%lu <- job item: %p data: %p thd: %p",
worker->id, job_item, ev, thd));
if (is_gtid_event(ev))
seen_gtid= true;
if (!seen_begin && ev->starts_group())
{
seen_begin= true; // The current group is started with B-event
worker->end_group_sets_max_dbs= true;
}
set_timespec_nsec(&worker->ts_exec[0], 0); // pre-exec
worker->stats_read_time += diff_timespec(&worker->ts_exec[0],
&worker->ts_exec[1]);
/* Adapting to possible new Format_description_log_event */
ptr_g= rli->gaq->get_job_group(ev->mts_group_idx);
if (ptr_g->new_fd_event)
{
worker->set_rli_description_event(ptr_g->new_fd_event);
ptr_g->new_fd_event= NULL;
}
error= worker->slave_worker_exec_event(ev);
set_timespec_nsec(&worker->ts_exec[1], 0); // pre-exec
worker->stats_exec_time += diff_timespec(&worker->ts_exec[1],
&worker->ts_exec[0]);
if (error || worker->found_order_commit_deadlock())
{
error= worker->retry_transaction(start_relay_number, start_relay_pos,
job_item->relay_number,
job_item->relay_pos);
if (error)
goto err;
}
/*
p-event or any other event of B-free (malformed) group can
"commit" with logical clock scheduler. In that case worker id
points to the only active "exclusive" Worker that processes such
malformed group events one by one.
WL#7592 refines the original assert disjunction formula
with the final disjunct.
*/
DBUG_ASSERT(seen_begin || is_gtid_event(ev) ||
ev->get_type_code() == binary_log::QUERY_EVENT ||
is_mts_db_partitioned(rli) || worker->id == 0 || seen_gtid);
if (ev->ends_group() ||
(!seen_begin && !is_gtid_event(ev) &&
(ev->get_type_code() == binary_log::QUERY_EVENT ||
/* break through by LC only in GTID off */
(!seen_gtid && !is_mts_db_partitioned(rli)))))
break;
remove_item_from_jobs(job_item, worker, rli);
/* The event will be used later if worker is NULL, so it is not freed */
if (ev->worker != NULL)
delete ev;
job_item= pop_jobs_item(worker, job_item);
}
DBUG_PRINT("info", (" commits GAQ index %lu, last committed %lu",
ev->mts_group_idx, worker->last_group_done_index));
/* The group is applied successfully, so error should be 0 */
worker->slave_worker_ends_group(ev, 0);
#ifndef DBUG_OFF
DBUG_PRINT("mts", ("Check_slave_debug_group worker %lu mts_checkpoint_group"
" %u processed %lu debug %d\n", worker->id, opt_mts_checkpoint_group,
worker->groups_done,
DBUG_EVALUATE_IF("check_slave_debug_group", 1, 0)));
if (DBUG_EVALUATE_IF("check_slave_debug_group", 1, 0) &&
opt_mts_checkpoint_group == worker->groups_done)
{
DBUG_PRINT("mts", ("Putting worker %lu in busy wait.", worker->id));
while (true) my_sleep(6000000);
}
#endif
remove_item_from_jobs(job_item, worker, rli);
delete ev;
DBUG_RETURN(0);
err:
if (error)
{
report_error_to_coordinator(worker);
DBUG_PRINT("info", ("Worker %lu is exiting: killed %i, error %i, "
"running_status %d",
worker->id, thd->killed, thd->is_error(),
worker->running_status));
worker->slave_worker_ends_group(ev, error); /* last done sets post exec */
}
DBUG_RETURN(error);
}
const char* Slave_worker::get_for_channel_str(bool upper_case) const
{
return c_rli->get_for_channel_str(upper_case);
}
const uint* Slave_worker::get_table_pk_field_indexes()
{
return info_slave_worker_table_pk_field_indexes;
}
uint Slave_worker::get_channel_field_index()
{
return LINE_FOR_CHANNEL;
}
Reference in New Issue View Git Blame Copy Permalink