Table of Contents
MySQL 8.0 includes the
sys schema, a set of objects that
helps DBAs and developers interpret data collected by the
Performance Schema. sys schema objects
can be used for typical tuning and diagnosis use cases. Objects in
this schema include:
Views that summarize Performance Schema data into more easily understandable form.
Stored procedures that perform operations such as Performance Schema configuration and generating diagnostic reports.
Stored functions that query Performance Schema configuration and provide formatting services.
For new installations, the sys schema
is installed by default during data directory initialization if you
use mysqld with the
--initialize or
--initialize-insecure option. If this
is not desired, you can drop the sys
schema manually after initialization if it is unneeded.
mysql_upgrade returns an error if a
sys schema exists but has no
version view, on the assumption that
absence of this view indicates a user-created sys
schema. To upgrade in this case, remove or rename the existing
sys schema first.
sys schema objects have a
DEFINER of
'mysql.sys'@'localhost'. Use of the dedicated
mysql.sys account avoids problems that occur if a
DBA renames or removes the root account.
Before using the sys schema, the
prerequisites described in this section must be satisfied.
Because the sys schema provides an
alternative means of accessing the Performance Schema, the
Performance Schema must be enabled for the
sys schema to work. See
Section 26.3, “Performance Schema Startup Configuration”.
For full access to the sys schema, a
user must have these privileges:
INSERT and
UPDATE for the
sys_config table, if changes are
to be made to it
Additional privileges for certain
sys schema stored procedures and
functions, as noted in their descriptions; for example, the
ps_setup_save() procedure
It is also necessary to have privileges for the objects underlying
the sys schema objects:
Certain Performance Schema instruments and consumers must be
enabled and (for instruments) timed to take full advantage of
sys schema capabilities:
All wait instruments
All stage instruments
All statement instruments
xxx_currentxxx_history_long
You can use the sys schema itself to
enable all of the additional instruments and consumers:
CALL sys.ps_setup_enable_instrument('wait');
CALL sys.ps_setup_enable_instrument('stage');
CALL sys.ps_setup_enable_instrument('statement');
CALL sys.ps_setup_enable_consumer('current');
CALL sys.ps_setup_enable_consumer('history_long');
For many uses of the sys schema, the default
Performance Schema is sufficient for data collection. Enabling
all the instruments and consumers just mentioned has a
performance impact, so it is preferable to enable only the
additional configuration you need. Also, remember that if you
enable additional configuration, you can easily restore the
default configuration like this:
CALL sys.ps_setup_reset_to_default(TRUE);
You can make the sys schema the
default schema so that references to its objects need not be
qualified with the schema name:
mysql>USE sys;Database changed mysql>SELECT * FROM version;+-------------+---------------+ | sys_version | mysql_version | +-------------+---------------+ | 2.0.0 | 8.0.13-debug | +-------------+---------------+
(The version view shows the
sys schema and MySQL server
versions.)
To access sys schema objects while a
different schema is the default (or simply to be explicit),
qualify object references with the schema name:
mysql> SELECT * FROM sys.version;
+-------------+---------------+
| sys_version | mysql_version |
+-------------+---------------+
| 2.0.0 | 8.0.13-debug |
+-------------+---------------+
The sys schema contains many views that
summarize Performance Schema tables in various ways. Most of these
views come in pairs, such that one member of the pair has the same
name as the other member, plus a x$ prefix. For
example, the host_summary_by_file_io
view summarizes file I/O grouped by host and displays latencies
converted from picoseconds to more readable values (with units);
mysql> SELECT * FROM sys.host_summary_by_file_io;
+------------+-------+------------+
| host | ios | io_latency |
+------------+-------+------------+
| localhost | 67570 | 5.38 s |
| background | 3468 | 4.18 s |
+------------+-------+------------+
The x$host_summary_by_file_io view
summarizes the same data but displays unformatted picosecond
latencies:
mysql> SELECT * FROM sys.x$host_summary_by_file_io;
+------------+-------+---------------+
| host | ios | io_latency |
+------------+-------+---------------+
| localhost | 67574 | 5380678125144 |
| background | 3474 | 4758696829416 |
+------------+-------+---------------+
The view without the x$ prefix is intended to
provide output that is more user friendly and easier for humans to
read. The view with the x$ prefix that displays
the same values in raw form is intended more for use with other
tools that perform their own processing on the data. For
additional information about the differences between
non-x$ and x$ views, see
Section 27.4.3, “sys Schema Views”.
To examine sys schema object
definitions, use the appropriate
SHOW statement or
INFORMATION_SCHEMA query. For
example, to examine the definitions of the
session view and
format_bytes() function, use these
statements:
mysql>SHOW CREATE VIEW sys.session;mysql>SHOW CREATE FUNCTION sys.format_bytes;
However, those statements display the definitions in relatively
unformatted form. To view object definitions with more readable
formatting, access the individual .sql files
available from the sys schema
development website at
https://github.com/mysql/mysql-sys.
Neither mysqldump nor
mysqlpump dump the
sys schema by default. To generate a
dump file, name the sys schema
explicitly on the command line using either of these commands:
mysqldump --databases --routines sys > sys_dump.sql mysqlpump sys > sys_dump.sql
To reinstall the schema from the dump file, use this command:
mysql < sys_dump.sql
The following sys schema views
provide progress reporting for long-running transactions:
processlist session x$processlist x$session
Assuming that the required instruments and consumers are enabled,
the progress column of these views shows the
percentage of work completed for stages that support progress
reporting.
Stage progress reporting requires that the
events_stages_current consumer be enabled, as
well as the instruments for which progress information is desired.
Instruments for these stages currently support progress reporting:
stage/sql/Copying to tmp table stage/innodb/alter table (end) stage/innodb/alter table (flush) stage/innodb/alter table (insert) stage/innodb/alter table (log apply index) stage/innodb/alter table (log apply table) stage/innodb/alter table (merge sort) stage/innodb/alter table (read PK and internal sort) stage/innodb/buffer pool load
For stages that do not support estimated and completed work
reporting, or if the required instruments or consumers are not
enabled, the progress column is
NULL.
The sys schema includes tables and
triggers, views, and stored procedures and functions. The
following sections provide details for each of these objects.
The following tables list sys
schema objects and provide a short description of each one.
Table 27.1 sys Schema Tables and Triggers
| Table or Trigger Name | Description |
|---|---|
sys_config |
sys schema configuration options |
sys_config_insert_set_user |
sys_config insert trigger |
sys_config_update_set_user |
sys_config update trigger |
Table 27.2 sys Schema Views
Table 27.3 sys Schema Stored Procedures
| Procedure Name | Description |
|---|---|
create_synonym_db() |
Create synonym for schema |
diagnostics() |
Collect system diagnostic information |
execute_prepared_stmt() |
Execute prepared statement |
ps_setup_disable_background_threads() |
Disable background thread instrumentation |
ps_setup_disable_consumer() |
Disable consumers |
ps_setup_disable_instrument() |
Disable instruments |
ps_setup_disable_thread() |
Disable instrumentation for thread |
ps_setup_enable_background_threads() |
Enable background thread instrumentation |
ps_setup_enable_consumer() |
Enable consumers |
ps_setup_enable_instrument() |
Enable instruments |
ps_setup_enable_thread() |
Enable instrumentation for thread |
ps_setup_reload_saved() |
Reload saved Performance Schema configuration |
ps_setup_reset_to_default() |
Reset saved Performance Schema configuration |
ps_setup_save() |
Save Performance Schema configuration |
ps_setup_show_disabled() |
Display disabled Performance Schema configuration |
ps_setup_show_disabled_consumers() |
Display disabled Performance Schema consumers |
ps_setup_show_disabled_instruments() |
Display disabled Performance Schema instruments |
ps_setup_show_enabled() |
Display enabled Performance Schema configuration |
ps_setup_show_enabled_consumers() |
Display enabled Performance Schema consumers |
ps_setup_show_enabled_instruments() |
Display enabled Performance Schema instruments |
ps_statement_avg_latency_histogram() |
Display statement latency histogram |
ps_trace_statement_digest() |
Trace Performance Schema instrumentation for digest |
ps_trace_thread() |
Dump Performance Schema data for thread |
ps_truncate_all_tables() |
Truncate Performance Schema summary tables |
statement_performance_analyzer() |
Report of statements running on server |
table_exists() |
Whether a table exists |
Table 27.4 sys Schema Stored Functions
| Function Name | Description |
|---|---|
extract_schema_from_file_name() |
Extract schema name from file path name |
extract_table_from_file_name() |
Extract table name from file path name |
format_bytes() |
Convert byte value to value with units |
format_path() |
Replace data and temp-file directories in path name with symbolic values |
format_statement() |
Truncate long statement to fixed length |
format_time() |
Convert picoseconds value to value with units |
list_add() |
Add item to list |
list_drop() |
Remove item from list |
ps_is_account_enabled() |
Check whether account instrumentation is enabled |
ps_is_consumer_enabled() |
Check whether consumer is enabled |
ps_is_instrument_default_enabled() |
Check whether instrument is enabled |
ps_is_instrument_default_timed() |
Check whether instrument is timed |
ps_is_thread_instrumented() |
Check whether thread is instrumented |
ps_thread_account() |
Return account for thread ID |
ps_thread_id() |
Return thread ID for connection ID |
ps_thread_stack() |
Return event information for thread ID |
ps_thread_trx_info() |
Return transaction information for thread ID |
quote_identifier() |
Return string as quoted identifier |
sys_get_config() |
Return sys schema configuration option |
version_major() |
MySQL server major version number |
version_minor() |
MySQL server minor version number |
version_patch() |
MySQL server patch release version number |
The following sections describe
sys schema tables and triggers.
This table contains sys schema
configuration options, one row per option. Configuration
changes made by updating this table persist across client
sessions and server restarts.
The sys_config table has these
columns:
variable
The configuration option name.
value
The configuration option value.
set_time
The timestamp of the most recent modification to the row.
set_by
The account that made the most recent modification to the
row. The value is NULL if the row has
not been changed since the
sys schema was installed.
As an efficiency measure to minimize the number of direct
reads from the sys_config table,
sys schema functions that use a
value from this table check for a user-defined variable with a
corresponding name, which is the user-defined variable having
the same name plus a @sys. prefix. (For
example, the variable corresponding to the
diagnostics.include_raw option is
@sys.diagnostics.include_raw.) If the
user-defined variable exists in the current session and is
non-NULL, the function uses its value in
preference to the value in the
sys_config table. Otherwise, the
function reads and uses the value from the table. In the
latter case, the calling function conventionally also sets the
corresponding user-defined variable to the table value so that
further references to the configuration option within the same
session use the variable and need not read the table again.
For example, the statement_truncate_len
option controls the maximum length of statements returned by
the format_statement() function.
The default is 64. To temporarily change the value to 32 for
your current session, set the corresponding
@sys.statement_truncate_len user-defined
variable:
mysql>SET @stmt = 'SELECT variable, value, set_time, set_by FROM sys_config';mysql>SELECT sys.format_statement(@stmt);+----------------------------------------------------------+ | sys.format_statement(@stmt) | +----------------------------------------------------------+ | SELECT variable, value, set_time, set_by FROM sys_config | +----------------------------------------------------------+ mysql>SET @sys.statement_truncate_len = 32;mysql>SELECT sys.format_statement(@stmt);+-----------------------------------+ | sys.format_statement(@stmt) | +-----------------------------------+ | SELECT variabl ... ROM sys_config | +-----------------------------------+
Subsequent invocations of
format_statement() within the
session continue to use the user-defined variable value (32),
rather than the value stored in the table (64).
To stop using the user-defined variable and revert to using
the value in the table, set the variable to
NULL within your session:
mysql>SET @sys.statement_truncate_len = NULL;mysql>SELECT sys.format_statement(@stmt);+----------------------------------------------------------+ | sys.format_statement(@stmt) | +----------------------------------------------------------+ | SELECT variable, value, set_time, set_by FROM sys_config | +----------------------------------------------------------+
Alternatively, end your current session (causing the user-defined variable to no longer exist) and begin a new session.
The conventional relationship just described between options
in the sys_config table and
user-defined variables can be exploited to make temporary
configuration changes that end when your session ends.
However, if you set a user-defined variable and then
subsequently change the corresponding table value within the
same session, the changed table value will not be used in that
session as long as the user-defined variable exists with a
non-NULL value. (The changed table value
will be used in other sessions that do
not have the user-defined variable assigned.)
The following list describes the options in the
sys_config table and the
corresponding user-defined variables:
diagnostics.allow_i_s_tables,
@sys.diagnostics.allow_i_s_tables
If this option is ON, the
diagnostics() procedure is
permitted to perform table scans on the
INFORMATION_SCHEMA.TABLES
table. This can be expensive if there are many tables. The
default is OFF.
diagnostics.include_raw,
@sys.diagnostics.include_raw
If this option is ON, the
diagnostics() procedure
includes the raw output from querying the
metrics view. The default is
OFF.
ps_thread_trx_info.max_length,
@sys.ps_thread_trx_info.max_length
The maximum length for JSON output produced by the
ps_thread_trx_info()
function. The default is 65535.
statement_performance_analyzer.limit,
@sys.statement_performance_analyzer.limit
The maximum number of rows to return for views that have
no built-in limit. (For example, the
statements_with_runtimes_in_95th_percentile
view has a built-in limit in the sense that it returns
only statements with average execution time in the 95th
percentile.) The default is 100.
statement_performance_analyzer.view,
@sys.statement_performance_analyzer.view
The custom query or view to be used by the
statement_performance_analyzer()
procedure (which is itself invoked by the
diagnostics() procedure). If
the option value contains a space, it is interpreted as a
query. Otherwise, it must be the name of an existing view
that queries the Performance Schema
events_statements_summary_by_digest
table. There cannot be any LIMIT clause
in the query or view definition if the
statement_performance_analyzer.limit
configuration option is greater than 0. The default is
NULL (no custom view defined).
statement_truncate_len,
@sys.statement_truncate_len
The maximum length of statements returned by the
format_statement() function.
Longer statements are truncated to this length. The
default is 64.
Other options can be added to the
sys_config table. For example,
the diagnostics() and
execute_prepared_stmt()
procedures use the debug option if it
exists, but this option is not part of the
sys_config table by default
because debug output normally is enabled only temporarily, by
setting the corresponding @sys.debug
user-defined variable. To enable debug output without having
to set that variable in individual sessions, add the option to
the table:
mysql> INSERT INTO sys.sys_config (variable, value) VALUES('debug', 'ON');
To change the debug setting in the table, do two things. First, modify the value in the table itself:
mysql>UPDATE sys.sys_configSET value = 'OFF'WHERE variable = 'debug';
Second, to also ensure that procedure invocations within the
current session use the changed value from the table, set the
corresponding user-defined variable to
NULL:
mysql> SET @sys.debug = NULL;
For rows added to the sys_config
table by INSERT statements, the
sys_config_insert_set_user
trigger sets the set_by column to the
current user.
The sys_config_update_set_user
trigger for the sys_config table
is similar to the
sys_config_insert_set_user
trigger, but for UPDATE
statements.
The following sections describe
sys schema views.
The sys schema contains many views that
summarize Performance Schema tables in various ways. Most of
these views come in pairs, such that one member of the pair has
the same name as the other member, plus a x$
prefix. For example, the
host_summary_by_file_io view
summarizes file I/O grouped by host and displays latencies
converted from picoseconds to more readable values (with units);
mysql> SELECT * FROM sys.host_summary_by_file_io;
+------------+-------+------------+
| host | ios | io_latency |
+------------+-------+------------+
| localhost | 67570 | 5.38 s |
| background | 3468 | 4.18 s |
+------------+-------+------------+
The x$host_summary_by_file_io view
summarizes the same data but displays unformatted picosecond
latencies:
mysql> SELECT * FROM sys.x$host_summary_by_file_io;
+------------+-------+---------------+
| host | ios | io_latency |
+------------+-------+---------------+
| localhost | 67574 | 5380678125144 |
| background | 3474 | 4758696829416 |
+------------+-------+---------------+
The view without the x$ prefix is intended to
provide output that is more user friendly and easier to read.
The view with the x$ prefix that displays the
same values in raw form is intended more for use with other
tools that perform their own processing on the data.
Views without the x$ prefix differ from the
corresponding x$ views in these ways:
Byte values are formatted with size units using
format_bytes().
Time values are formatted with temporal units using
format_time().
SQL statements are truncated to a maximum display width
using format_statement().
Path name are shortened using
format_path().
These views summarize statement activity, file I/O, and connections, grouped by host.
The host_summary and
x$host_summary views have these
columns:
host
The host from which the client connected. Rows for which
the HOST column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
statements
The total number of statements for the host.
statement_latency
The total wait time of timed statements for the host.
statement_avg_latency
The average wait time per timed statement for the host.
table_scans
The total number of table scans for the host.
file_ios
The total number of file I/O events for the host.
file_io_latency
The total wait time of timed file I/O events for the host.
current_connections
The current number of connections for the host.
total_connections
The total number of connections for the host.
unique_users
The number of distinct users for the host.
current_memory
The current amount of allocated memory for the host.
total_memory_allocated
The total amount of allocated memory for the host.
These views summarize file I/O, grouped by host. By default, rows are sorted by descending total file I/O latency.
The host_summary_by_file_io and
x$host_summary_by_file_io views
have these columns:
host
The host from which the client connected. Rows for which
the HOST column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
ios
The total number of file I/O events for the host.
io_latency
The total wait time of timed file I/O events for the host.
These views summarize file I/O, grouped by host and event type. By default, rows are sorted by host and descending total I/O latency.
The host_summary_by_file_io_type
and
x$host_summary_by_file_io_type
views have these columns:
host
The host from which the client connected. Rows for which
the HOST column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
event_name
The file I/O event name.
total
The total number of occurrences of the file I/O event for the host.
total_latency
The total wait time of timed occurrences of the file I/O event for the host.
max_latency
The maximum single wait time of timed occurrences of the file I/O event for the host.
These views summarize statement stages, grouped by host. By default, rows are sorted by host and descending total latency.
The host_summary_by_stages and
x$host_summary_by_stages views
have these columns:
host
The host from which the client connected. Rows for which
the HOST column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
event_name
The stage event name.
total
The total number of occurrences of the stage event for the host.
total_latency
The total wait time of timed occurrences of the stage event for the host.
avg_latency
The average wait time per timed occurrence of the stage event for the host.
These views summarize overall statement statistics, grouped by host. By default, rows are sorted by descending total latency.
The
host_summary_by_statement_latency
and
x$host_summary_by_statement_latency
views have these columns:
host
The host from which the client connected. Rows for which
the HOST column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
total
The total number of statements for the host.
total_latency
The total wait time of timed statements for the host.
max_latency
The maximum single wait time of timed statements for the host.
lock_latency
The total time waiting for locks by timed statements for the host.
rows_sent
The total number of rows returned by statements for the host.
rows_examined
The total number of rows read from storage engines by statements for the host.
rows_affected
The total number of rows affected by statements for the host.
full_scans
The total number of full table scans by statements for the host.
These views summarize informaion about statements executed, grouped by host and statement type. By default, rows are sorted by host and descending total latency.
The
host_summary_by_statement_type
and
x$host_summary_by_statement_type
views have these columns:
host
The host from which the client connected. Rows for which
the HOST column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
statement
The final component of the statement event name.
total
The total number of occurrences of the statement event for the host.
total_latency
The total wait time of timed occurrences of the statement event for the host.
max_latency
The maximum single wait time of timed occurrences of the statement event for the host.
lock_latency
The total time waiting for locks by timed occurrences of the statement event for the host.
rows_sent
The total number of rows returned by occurrences of the statement event for the host.
rows_examined
The total number of rows read from storage engines by occurrences of the statement event for the host.
rows_affected
The total number of rows affected by occurrences of the statement event for the host.
full_scans
The total number of full table scans by occurrences of the statement event for the host.
These views summarize the information in the
INFORMATION_SCHEMA
INNODB_BUFFER_PAGE table, grouped
by schema. By default, rows are sorted by descending buffer
size.
Querying views that access the
INNODB_BUFFER_PAGE table can
affect performance. Do not query these views on a production
system unless you are aware of the performance impact and
have determined it to be acceptable. To avoid impacting
performance on a production system, reproduce the issue you
want to investigate and query buffer pool statistics on a
test instance.
The
innodb_buffer_stats_by_schema
and
x$innodb_buffer_stats_by_schema
views have these columns:
object_schema
The schema name for the object, or InnoDB
System if the table belongs to the
InnoDB storage engine.
allocated
The total number of bytes allocated for the schema.
data
The total number of data bytes allocated for the schema.
pages
The total number of pages allocated for the schema.
pages_hashed
The total number of hashed pages allocated for the schema.
pages_old
The total number of old pages allocated for the schema.
rows_cached
The total number of cached rows for the schema.
These views summarize the information in the
INFORMATION_SCHEMA
INNODB_BUFFER_PAGE table, grouped
by schema and table. By default, rows are sorted by descending
buffer size.
Querying views that access the
INNODB_BUFFER_PAGE table can
affect performance. Do not query these views on a production
system unless you are aware of the performance impact and
have determined it to be acceptable. To avoid impacting
performance on a production system, reproduce the issue you
want to investigate and query buffer pool statistics on a
test instance.
The innodb_buffer_stats_by_table
and
x$innodb_buffer_stats_by_table
views have these columns:
object_schema
The schema name for the object, or InnoDB
System if the table belongs to the
InnoDB storage engine.
object_name
The table name.
allocated
The total number of bytes allocated for the table.
data
The number of data bytes allocated for the table.
pages
The total number of pages allocated for the table.
pages_hashed
The number of hashed pages allocated for the table.
pages_old
The number of old pages allocated for the table.
rows_cached
The number of cached rows for the table.
These views summarize the InnoDB locks that
transactions are waiting for. By default, rows are sorted by
descending lock age.
The innodb_lock_waits and
x$innodb_lock_waits views have
these columns:
wait_started
The time at which the lock wait started.
wait_age
How long the lock has been waited for, as a
TIME value.
wait_age_secs
How long the lock has been waited for, in seconds.
locked_table_schema
The schema that contains the locked table.
locked_table_name
The name of the locked table.
locked_table_partition
The name of the locked partition, if any;
NULL otherwise.
locked_table_subpartition
The name of the locked subpartition, if any;
NULL otherwise.
locked_index
The name of the locked index.
locked_type
The type of the waiting lock.
waiting_trx_id
The ID of the waiting transaction.
waiting_trx_started
The time at which the waiting transaction started.
waiting_trx_age
How long the waiting transaction has been waiting, as a
TIME value.
waiting_trx_rows_locked
The number of rows locked by the waiting transaction.
waiting_trx_rows_modified
The number of rows modified by the waiting transaction.
waiting_pid
The processlist ID of the waiting transaction.
waiting_query
The statement that is waiting for the lock.
waiting_lock_id
The ID of the waiting lock.
waiting_lock_mode
The mode of the waiting lock.
blocking_trx_id
The ID of the transaction that is blocking the waiting lock.
blocking_pid
The processlist ID of the blocking transaction.
blocking_query
The statement the blocking transaction is executing. This field reports NULL if the session that issued the blocking query becomes idle. For more information, see Identifying a Blocking Query After the Issuing Session Becomes Idle.
blocking_lock_id
The ID of the lock that is blocking the waiting lock.
blocking_lock_mode
The mode of the lock that is blocking the waiting lock.
blocking_trx_started
The time at which the blocking transaction started.
blocking_trx_age
How long the blocking transaction has been executing, as a
TIME value.
blocking_trx_rows_locked
The number of rows locked by the blocking transaction.
blocking_trx_rows_modified
The number of rows modified by the blocking transaction.
sql_kill_blocking_query
The KILL statement to
execute to kill the blocking statement.
sql_kill_blocking_connection
The KILL statement to
execute to kill the session running the blocking
statement.
These views summarize I/O consumers to display time waiting for I/O, grouped by thread. By default, rows are sorted by descending total I/O latency.
The io_by_thread_by_latency and
x$io_by_thread_by_latency views
have these columns:
user
For foreground threads, the account associated with the thread. For background threads, the thread name.
total
The total number of I/O events for the thread.
total_latency
The total wait time of timed I/O events for the thread.
min_latency
The minimum single wait time of timed I/O events for the thread.
avg_latency
The average wait time per timed I/O event for the thread.
max_latency
The maximum single wait time of timed I/O events for the thread.
thread_id
The thread ID.
processlist_id
For foreground threads, the processlist ID of the thread.
For background threads, NULL.
These views summarize global I/O consumers to display amount of I/O, grouped by file. By default, rows are sorted by descending total I/O (bytes read and written).
The io_global_by_file_by_bytes
and x$io_global_by_file_by_bytes
views have these columns:
file
The file path name.
count_read
The total number of read events for the file.
total_read
The total number of bytes read from the file.
avg_read
The average number of bytes per read from the file.
count_write
The total number of write events for the file.
total_written
The total number of bytes written to the file.
avg_write
The average number of bytes per write to the file.
total
The total number of bytes read and written for the file.
write_pct
The percentage of total bytes of I/O that were writes.
These views summarize global I/O consumers to display time waiting for I/O, grouped by file. By default, rows are sorted by descending total latency.
The io_global_by_file_by_latency
and
x$io_global_by_file_by_latency
views have these columns:
file
The file path name.
total
The total number of I/O events for the file.
total_latency
The total wait time of timed I/O events for the file.
count_read
The total number of read I/O events for the file.
read_latency
The total wait time of timed read I/O events for the file.
count_write
The total number of write I/O events for the file.
write_latency
The total wait time of timed write I/O events for the file.
count_misc
The total number of other I/O events for the file.
misc_latency
The total wait time of timed other I/O events for the file.
These views summarize global I/O consumers to display amount of I/O and time waiting for I/O, grouped by event. By default, rows are sorted by descending total I/O (bytes read and written).
The io_global_by_wait_by_bytes
and x$io_global_by_wait_by_bytes
views have these columns:
event_name
The I/O event name, with the
wait/io/file/ prefix stripped.
total
The total number of occurrences of the I/O event.
total_latency
The total wait time of timed occurrences of the I/O event.
min_latency
The minimum single wait time of timed occurrences of the I/O event.
avg_latency
The average wait time per timed occurrence of the I/O event.
max_latency
The maximum single wait time of timed occurrences of the I/O event.
count_read
The number of read requests for the I/O event.
total_read
The number of bytes read for the I/O event.
avg_read
The average number of bytes per read for the I/O event.
count_write
The number of write requests for the I/O event.
total_written
The number of bytes written for the I/O event.
avg_written
The average number of bytes per write for the I/O event.
total_requested
The total number of bytes read and written for the I/O event.
These views summarize global I/O consumers to display amount of I/O and time waiting for I/O, grouped by event. By default, rows are sorted by descending total latency.
The io_global_by_wait_by_latency
and
x$io_global_by_wait_by_latency
views have these columns:
event_name
The I/O event name, with the
wait/io/file/ prefix stripped.
total
The total number of occurrences of the I/O event.
total_latency
The total wait time of timed occurrences of the I/O event.
avg_latency
The average wait time per timed occurrence of the I/O event.
max_latency
The maximum single wait time of timed occurrences of the I/O event.
read_latency
The total wait time of timed read occurrences of the I/O event.
write_latency
The total wait time of timed write occurrences of the I/O event.
misc_latency
The total wait time of timed other occurrences of the I/O event.
count_read
The number of read requests for the I/O event.
total_read
The number of bytes read for the I/O event.
avg_read
The average number of bytes per read for the I/O event.
count_write
The number of write requests for the I/O event.
total_written
The number of bytes written for the I/O event.
avg_written
The average number of bytes per write for the I/O event.
These views summarize file I/O activity, grouped by file and thread. By default, rows are sorted with most recent I/O first.
The latest_file_io and
x$latest_file_io views have
these columns:
thread
For foreground threads, the account associated with the thread (and port number for TCP/IP connections). For background threads, the thread name and thread ID
file
The file path name.
latency
The wait time of the file I/O event.
operation
The type of operation.
requested
The number of data bytes requested for the file I/O event.
These views summarize memory use, grouped by host. By default, rows are sorted by descending amount of memory used.
The
memory_by_host_by_current_bytes
and
x$memory_by_host_by_current_bytes
views have these columns:
host
The host from which the client connected. Rows for which
the HOST column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
current_count_used
The current number of allocated memory blocks that have not been freed yet for the host.
current_allocated
The current number of allocated bytes that have not been freed yet for the host.
current_avg_alloc
The current number of allocated bytes per memory block for the host.
current_max_alloc
The largest single current memory allocation in bytes for the host.
total_allocated
The total memory allocation in bytes for the host.
These views summarize memory use, grouped by thread. By default, rows are sorted by descending amount of memory used.
The
memory_by_thread_by_current_bytes
and
x$memory_by_thread_by_current_bytes
views have these columns:
thread_id
The thread ID.
user
The thread user or thread name.
current_count_used
The current number of allocated memory blocks that have not been freed yet for the thread.
current_allocated
The current number of allocated bytes that have not been freed yet for the thread.
current_avg_alloc
The current number of allocated bytes per memory block for the thread.
current_max_alloc
The largest single current memory allocation in bytes for the thread.
total_allocated
The total memory allocation in bytes for the thread.
These views summarize memory use, grouped by user. By default, rows are sorted by descending amount of memory used.
The
memory_by_user_by_current_bytes
and
x$memory_by_user_by_current_bytes
views have these columns:
user
The client user name. Rows for which the
USER column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
current_count_used
The current number of allocated memory blocks that have not been freed yet for the user.
current_allocated
The current number of allocated bytes that have not been freed yet for the user.
current_avg_alloc
The current number of allocated bytes per memory block for the user.
current_max_alloc
The largest single current memory allocation in bytes for the user.
total_allocated
The total memory allocation in bytes for the user.
These views summarize memory use, grouped by allocation type (that is, by event). By default, rows are sorted by descending amount of memory used.
The
memory_global_by_current_bytes
and
x$memory_global_by_current_bytes
views have these columns:
event_name
The memory event name.
current_count
The total number of occurrences of the event.
current_alloc
The current number of allocated bytes that have not been freed yet for the event.
current_avg_alloc
The current number of allocated bytes per memory block for the event.
high_count
The high-water mark for number of memory blocks allocated for the event.
high_alloc
The high-water mark for number of bytes allocated for the event.
high_avg_alloc
The high-water mark for average number of bytes per memory block allocated for the event.
These views summarize total memory use within the server.
The memory_global_total and
x$memory_global_total views have
these columns:
total_allocated
The total bytes of memory allocated within the server.
This view summarizes MySQL server metrics to show variable names, values, types, and whether they are enabled. By default, rows are sorted by variable type and name.
The metrics view includes this
information:
Global status variables from the Performance Schema
global_status table
InnoDB metrics from the
INFORMATION_SCHEMA
INNODB_METRICS table
Current and total memory allocation, based on the Performance Schema memory instrumentation
The current time (human readable and Unix timestamp formats)
There is some duplication of information between the
global_status and
INNODB_METRICS tables, which the
metrics view eliminates.
The metrics view has these
columns:
Variable_name
The metric name. The metric type determines the source from which the name is taken:
For global status variables: The
VARIABLE_NAME column of the
global_status table
For InnoDB metrics: The
NAME column of the
INNODB_METRICS table
For other metrics: A view-provided descriptive string
Variable_value
The metric value. The metric type determines the source from which the value is taken:
For global status variables: The
VARIABLE_VALUE column of the
global_status table
For InnoDB metrics: The
COUNT column of the
INNODB_METRICS table
For memory metrics: The relevant column from the
Performance Schema
memory_summary_global_by_event_name
table
For the current time: The value of
NOW(3) or
UNIX_TIMESTAMP(NOW(3))
Type
The metric type:
For global status variables: Global
Status
For InnoDB metrics: InnoDB
Metrics - %, where % is
replaced by the value of the
SUBSYSTEM column of the
INNODB_METRICS table
For memory metrics: Performance
Schema
For the current time: System Time
Enabled
Whether the metric is enabled:
For global status variables: YES
For InnoDB metrics:
YES if the
STATUS column of the
INNODB_METRICS table is
enabled, NO
otherwise
For memory metrics: NO,
YES, or PARTIAL
(currently, PARTIAL occurs only for
memory metrics and indicates that not all
memory/% instruments are enabled;
Performance Schema memory instruments are always
enabled)
For the current time: YES
These views summarize processlist information. They provide
more complete information than the SHOW
PROCESSLIST statement and the
INFORMATION_SCHEMA
PROCESSLIST table, and are also
nonblocking. By default, rows are sorted by descending process
time and descending wait time.
The column descriptions here are brief. For additional
information, see the description of the Performance Schema
threads table at
Section 26.12.17.5, “The threads Table”.
The processlist and
x$processlist views have these
columns:
thd_id
The thread ID.
conn_id
The connection ID.
user
The thread user or thread name.
db
The default database for the thread, or
NULL if there is none.
command
For foreground threads, the type of command the thread is
executing on behalf of the client, or
Sleep if the session is idle.
state
An action, event, or state that indicates what the thread is doing.
time
The time in seconds that the thread has been in its current state.
current_statement
The statement the thread is executing, or
NULL if it is not executing any
statement.
statement_latency
How long the statement has been executing.
progress
The percentage of work completed for stages that support progress reporting. See Section 27.3, “sys Schema Progress Reporting”.
lock_latency
The time spent waiting for locks by the current statement.
rows_examined
The number of rows read from storage engines by the current statement.
rows_sent
The number of rows returned by the current statement.
rows_affected
The number of rows affected by the current statement.
tmp_tables
The number of internal in-memory temporary tables created by the current statement.
tmp_disk_tables
The number of internal on-disk temporary tables created by the current statement.
full_scan
The number of full table scans performed by the current statement.
last_statement
The last statement executed by the thread, if there is no currently executing statement or wait.
last_statement_latency
How long the last statement executed.
current_memory
The number of bytes allocated by the thread.
last_wait
The name of the most recent wait event for the thread.
last_wait_latency
The wait time of the most recent wait event for the thread.
source
The source file and line number containing the instrumented code that produced the event.
trx_latency
The wait time of the current transaction for the thread.
trx_state
The state for the current transaction for the thread.
trx_autocommit
Whether autocommit mode was enabled when the current transaction started.
pid
The client process ID.
program_name
The client program name.
This view returns information about lost Performance Schema instruments, to indicate whether the Performance Schema is unable to monitor all runtime data.
The
ps_check_lost_instrumentation
view has these columns:
variable_name
The Performance Schema status variable name indicating which type of instrument was lost.
variable_value
The number of instruments lost.
This view indicates which tables have
AUTO_INCREMENT columns and provides
information about those columns, such as the current and
maximum column values and the usage ratio (ratio of used to
possible values). By default, rows are sorted by descending
usage ratio and maximum column value.
Tables in these schemas are excluded from view output:
mysql, sys,
INFORMATION_SCHEMA,
performance_schema.
The
schema_auto_increment_columns
view has these columns:
table_schema
The schema that contains the table.
table_name
The table that contains the
AUTO_INCREMENT column.
column_name
The name of the AUTO_INCREMENT column.
data_type
The data type of the column.
column_type
The column type of the column, which is the data type plus
possibly other information. For example, for a column with
a bigint(20) unsigned column type, the
data type is just bigint.
is_signed
Whether the column type is signed.
is_unsigned
Whether the column type is unsigned.
max_value
The maximum permitted value for the column.
auto_increment
The current AUTO_INCREMENT value for
the column.
auto_increment_ratio
The ratio of used to permitted values for the column. This indicates how much of the sequence of values is “used up.”
These views provide index statistics. By default, rows are sorted by descending total index latency.
The schema_index_statistics and
x$schema_index_statistics views
have these columns:
table_schema
The schema that contains the table.
table_name
The table that contains the index.
index_name
The name of the index.
rows_selected
The total number of rows read using the index.
select_latency
The total wait time of timed reads using the index.
rows_inserted
The total number of rows inserted into the index.
insert_latency
The total wait time of timed inserts into the index.
rows_updated
The total number of rows updated in the index.
update_latency
The total wait time of timed updates in the index.
rows_deleted
The total number of rows deleted from the index.
delete_latency
The total wait time of timed deletes from the index.
This view summarizes the types of objects within each schema. By default, rows are sorted by schema and object type.
For MySQL instances with a large number of objects, this view might take a long time to execute.
The schema_object_overview view
has these columns:
db
The schema name.
object_type
The object type: BASE TABLE,
INDEX
(,
index_type)EVENT, FUNCTION,
PROCEDURE, TRIGGER,
VIEW.
count
The number of objects in the schema of the given type.
The schema_redundant_indexes
view displays indexes that duplicate other indexes or are made
redundant by them. The
x$schema_flattened_keys
view is a helper view for
schema_redundant_indexes.
In the following column descriptions, the dominant index is the one that makes the redundant index redundant.
The schema_redundant_indexes
view has these columns:
table_schema
The schema that contains the table.
table_name
The table that contains the index.
redundant_index_name
The name of the redundant index.
redundant_index_columns
The names of the columns in the redundant index.
redundant_index_non_unique
The number of nonunique columns in the redundant index.
dominant_index_name
The name of the dominant index.
dominant_index_columns
The names of the columns in the dominant index.
dominant_index_non_unique
The number of nonunique columns in the dominant index.
subpart_exists
Whether the index indexes only part of a column.
sql_drop_index
The statement to execute to drop the redundant index.
The
x$schema_flattened_keys
view has these columns:
table_schema
The schema that contains the table.
table_name
The table that contains the index.
index_name
An index name.
non_unique
The number of nonunique columns in the index.
subpart_exists
Whether the index indexes only part of a column.
index_columns
The name of the columns in the index.
These views display which sessions are blocked waiting on metadata locks, and what is blocking them.
The column descriptions here are brief. For additional
information, see the description of the Performance Schema
metadata_locks table at
Section 26.12.12.3, “The metadata_locks Table”.
The schema_table_lock_waits and
x$schema_table_lock_waits views
have these columns:
object_schema
The schema containing the object to be locked.
object_name
The name of the instrumented object.
waiting_thread_id
The thread ID of the thread that is waiting for the lock.
waiting_pid
The processlist ID of the thread that is waiting for the lock.
waiting_account
The account associated with the session that is waiting for the lock.
waiting_lock_type
The type of the waiting lock.
waiting_lock_duration
How long the waiting lock has been waiting.
waiting_query
The statement that is waiting for the lock.
waiting_query_secs
How long the statement has been waiting, in seconds.
waiting_query_rows_affected
The number of rows affected by the statement.
waiting_query_rows_examined
The number of rows read from storage engines by the statement.
blocking_thread_id
The thread ID of the thread that is blocking the waiting lock.
blocking_pid
The processlist ID of the thread that is blocking the waiting lock.
blocking_account
The account associated with the thread that is blocking the waiting lock.
blocking_lock_type
The type of lock that is blocking the waiting lock.
blocking_lock_duration
How long the blocking lock has been held.
sql_kill_blocking_query
The KILL statement to
execute to kill the blocking statement.
sql_kill_blocking_connection
The KILL statement to
execute to kill the session running the blocking
statement.
These views summarize table statistics. By default, rows are sorted by descending total wait time (tables with most contention first).
These views user a helper view,
x$ps_schema_table_statistics_io.
The schema_table_statistics and
x$schema_table_statistics views
have these columns:
table_schema
The schema that contains the table.
table_name
The table name.
total_latency
The total wait time of timed I/O events for the table.
rows_fetched
The total number of rows read from the table.
fetch_latency
The total wait time of timed read I/O events for the table.
rows_inserted
The total number of rows inserted into the table.
insert_latency
The total wait time of timed insert I/O events for the table.
rows_updated
The total number of rows updated in the table.
update_latency
The total wait time of timed update I/O events for the table.
rows_deleted
The total number of rows deleted from the table.
delete_latency
The total wait time of timed delete I/O events for the table.
io_read_requests
The total number of read requests for the table.
io_read
The total number of bytes read from the table.
io_read_latency
The total wait time of reads from the table.
io_write_requests
The total number of write requests for the table.
io_write
The total number of bytes written to the table.
io_write_latency
The total wait time of writes to the table.
io_misc_requests
The total number of miscellaneous I/O requests for the table.
io_misc_latency
The total wait time of miscellaneous I/O requests for the table.
These views summarize table statistics, including
InnoDB buffer pool statistics. By default,
rows are sorted by descending total wait time (tables with
most contention first).
These views user a helper view,
x$ps_schema_table_statistics_io.
The
schema_table_statistics_with_buffer
and
x$schema_table_statistics_with_buffer
views have these columns:
table_schema
The schema that contains the table.
table_name
The table name.
rows_fetched
The total number of rows read from the table.
fetch_latency
The total wait time of timed read I/O events for the table.
rows_inserted
The total number of rows inserted into the table.
insert_latency
The total wait time of timed insert I/O events for the table.
rows_updated
The total number of rows updated in the table.
update_latency
The total wait time of timed update I/O events for the table.
rows_deleted
The total number of rows deleted from the table.
delete_latency
The total wait time of timed delete I/O events for the table.
io_read_requests
The total number of read requests for the table.
io_read
The total number of bytes read from the table.
io_read_latency
The total wait time of reads from the table.
io_write_requests
The total number of write requests for the table.
io_write
The total number of bytes written to the table.
io_write_latency
The total wait time of writes to the table.
io_misc_requests
The total number of miscellaneous I/O requests for the table.
io_misc_latency
The total wait time of miscellaneous I/O requests for the table.
innodb_buffer_allocated
The total number of InnoDB buffer bytes
allocated for the table.
innodb_buffer_data
The total number of InnoDB data bytes
allocated for the table.
innodb_buffer_free
The total number of InnoDB nondata
bytes allocated for the table
(innodb_buffer_allocated −
innodb_buffer_data).
innodb_buffer_pages
The total number of InnoDB pages
allocated for the table.
innodb_buffer_pages_hashed
The total number of InnoDB hashed pages
allocated for the table.
innodb_buffer_pages_old
The total number of InnoDB old pages
allocated for the table.
innodb_buffer_rows_cached
The total number of InnoDB cached rows
for the table.
These views display which tables are being accessed with full table scans. By default, rows are sorted by descending rows scanned.
The
schema_tables_with_full_table_scans
and
x$schema_tables_with_full_table_scans
views have these columns:
object_schema
The schema name.
object_name
The table name.
rows_full_scanned
The total number of rows scanned by full scans of the table.
latency
The total wait time of full scans of the table.
These views display indexes for which there are no events, which indicates that they are not being used. By default, rows are sorted by schema and table.
This view is most useful when the server has been up and processing long enough that its workload is representative. Otherwise, presence of an index in this view may not be meaningful.
The schema_unused_indexes view
has these columns:
object_schema
The schema name.
object_name
The table name.
index_name
The unused index name.
These views are similar to
processlist and
x$processlist, but they filter
out background processes to display only user sessions. For
descriptions of the columns, see
Section 27.4.3.22, “The processlist and x$processlist Views”.
For each connection, this view displays the SSL version, cipher, and count of reused SSL sessions.
The session_ssl_status view has
these columns:
thread_id
The thread ID for the connection.
ssl_version
The version of SSL used for the connection.
ssl_cipher
The SSL cipher used for the connection.
ssl_sessions_reused
The number of reused SSL sessions for the connection.
These views list normalized statements with aggregated statistics. The content mimics the MySQL Enterprise Monitor Query Analysis view. By default, rows are sorted by descending total latency.
The statement_analysis and
x$statement_analysis views have
these columns:
query
The normalized statement string.
db
The default database for the statement, or
NULL if there is none.
full_scan
The total number of full table scans performed by occurrences of the statement.
exec_count
The total number of times the statement has executed.
err_count
The total number of errors produced by occurrences of the statement.
warn_count
The total number of warnings produced by occurrences of the statement.
total_latency
The total wait time of timed occurrences of the statement.
max_latency
The maximum single wait time of timed occurrences of the statement.
avg_latency
The average wait time per timed occurrence of the statement.
lock_latency
The total time waiting for locks by timed occurrences of the statement.
rows_sent
The total number of rows returned by occurrences of the statement.
rows_sent_avg
The average number of rows returned per occurrence of the statement.
rows_examined
The total number of rows read from storage engines by occurrences of the statement.
rows_examined_avg
The average number of rows read from storage engines per occurrence of the statement.
rows_affected
The total number of rows affected by occurrences of the statement.
rows_affected_avg
The average number of rows affected per occurrence of the statement.
tmp_tables
The total number of internal in-memory temporary tables created by occurrences of the statement.
tmp_disk_tables
The total number of internal on-disk temporary tables created by occurrences of the statement.
rows_sorted
The total number of rows sorted by occurrences of the statement.
sort_merge_passes
The total number of sort merge passes by occurrences of the statement.
digest
The statement digest.
first_seen
The time at which the statement was first seen.
last_seen
The time at which the statement was most recently seen.
These views display normalized statements that have produced errors or warnings. By default, rows are sorted by descending error and warning counts.
The
statements_with_errors_or_warnings
and
x$statements_with_errors_or_warnings
views have these columns:
query
The normalized statement string.
db
The default database for the statement, or
NULL if there is none.
exec_count
The total number of times the statement has executed.
errors
The total number of errors produced by occurrences of the statement.
error_pct
The percentage of statement occurrences that produced errors.
warnings
The total number of warnings produced by occurrences of the statement.
warning_pct
The percentage of statement occurrences that produced warnings.
first_seen
The time at which the statement was first seen.
last_seen
The time at which the statement was most recently seen.
digest
The statement digest.
These views display normalized statements that have done full table scans. By default, rows are sorted by descending percentage of time a full scan was done and descending total latency.
The
statements_with_full_table_scans
and
x$statements_with_full_table_scans
views have these columns:
query
The normalized statement string.
db
The default database for the statement, or
NULL if there is none.
exec_count
The total number of times the statement has executed.
total_latency
The total wait time of timed statement events for the statement.
no_index_used_count
The total number of times no index was used to scan the table.
no_good_index_used_count
The total number of times no good index was used to scan the table.
no_index_used_pct
The percentage of the time no index was used to scan the table.
rows_sent
The total number of rows returned from the table.
rows_examined
The total number of rows read from the storage engine for the table.
rows_sent_avg
The average number of rows returned from the table.
rows_examined_avg
The average number of rows read from the storage engine for the table.
first_seen
The time at which the statement was first seen.
last_seen
The time at which the statement was most recently seen.
digest
The statement digest.
These views list statements with runtimes in the 95th percentile. By default, rows are sorted by descending average latency.
Both views use two helper views,
x$ps_digest_avg_latency_distribution and
x$ps_digest_95th_percentile_by_avg_us.
The
statements_with_runtimes_in_95th_percentile
and
x$statements_with_runtimes_in_95th_percentile
views have these columns:
query
The normalized statement string.
db
The default database for the statement, or
NULL if there is none.
full_scan
The total number of full table scans performed by occurrences of the statement.
exec_count
The total number of times the statement has executed.
err_count
The total number of errors produced by occurrences of the statement.
warn_count
The total number of warnings produced by occurrences of the statement.
total_latency
The total wait time of timed occurrences of the statement.
max_latency
The maximum single wait time of timed occurrences of the statement.
avg_latency
The average wait time per timed occurrence of the statement.
rows_sent
The total number of rows returned by occurrences of the statement.
rows_sent_avg
The average number of rows returned per occurrence of the statement.
rows_examined
The total number of rows read from storage engines by occurrences of the statement.
rows_examined_avg
The average number of rows read from storage engines per occurrence of the statement.
first_seen
The time at which the statement was first seen.
last_seen
The time at which the statement was most recently seen.
digest
The statement digest.
These views list normalized statements that have performed sorts. By default, rows are sorted by descending total latency.
The statements_with_sorting and
x$statements_with_sorting views
have these columns:
query
The normalized statement string.
db
The default database for the statement, or
NULL if there is none.
exec_count
The total number of times the statement has executed.
total_latency
The total wait time of timed occurrences of the statement.
sort_merge_passes
The total number of sort merge passes by occurrences of the statement.
avg_sort_merges
The average number of sort merge passes per occurrence of the statement.
sorts_using_scans
The total number of sorts using table scans by occurrences of the statement.
sort_using_range
The total number of sorts using range accesses by occurrences of the statement.
rows_sorted
The total number of rows sorted by occurrences of the statement.
avg_rows_sorted
The average number of rows sorted per occurrence of the statement.
first_seen
The time at which the statement was first seen.
last_seen
The time at which the statement was most recently seen.
digest
The statement digest.
These views list normalized statements that have used temporary tables. By default, rows are sorted by descending number of on-disk temporary tables used and descending number of in-memory temporary tables used.
The statements_with_temp_tables
and
x$statements_with_temp_tables
views have these columns:
query
The normalized statement string.
db
The default database for the statement, or
NULL if there is none.
exec_count
The total number of times the statement has executed.
total_latency
The total wait time of timed occurrences of the statement.
memory_tmp_tables
The total number of internal in-memory temporary tables created by occurrences of the statement.
disk_tmp_tables
The total number of internal on-disk temporary tables created by occurrences of the statement.
avg_tmp_tables_per_query
The average number of internal temporary tables created per occurrence of the statement.
tmp_tables_to_disk_pct
The percentage of internal in-memory temporary tables that were converted to on-disk tables.
first_seen
The time at which the statement was first seen.
last_seen
The time at which the statement was most recently seen.
digest
The statement digest.
These views summarize statement activity, file I/O, and connections, grouped by user. By default, rows are sorted by descending total latency.
The user_summary and
x$user_summary views have these
columns:
user
The client user name. Rows for which the
USER column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
statements
The total number of statements for the user.
statement_latency
The total wait time of timed statements for the user.
statement_avg_latency
The average wait time per timed statement for the user.
table_scans
The total number of table scans for the user.
file_ios
The total number of file I/O events for the user.
file_io_latency
The total wait time of timed file I/O events for the user.
current_connections
The current number of connections for the user.
total_connections
The total number of connections for the user.
unique_hosts
The number of distinct hosts from which connections for the user have originated.
current_memory
The current amount of allocated memory for the user.
total_memory_allocated
The total amount of allocated memory for the user.
These views summarize file I/O, grouped by user. By default, rows are sorted by descending total file I/O latency.
The user_summary_by_file_io and
x$user_summary_by_file_io views
have these columns:
user
The client user name. Rows for which the
USER column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
ios
The total number of file I/O events for the user.
io_latency
The total wait time of timed file I/O events for the user.
These views summarize file I/O, grouped by user and event type. By default, rows are sorted by user and descending total latency.
The user_summary_by_file_io_type
and
x$user_summary_by_file_io_type
views have these columns:
user
The client user name. Rows for which the
USER column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
event_name
The file I/O event name.
total
The total number of occurrences of the file I/O event for the user.
latency
The total wait time of timed occurrences of the file I/O event for the user.
max_latency
The maximum single wait time of timed occurrences of the file I/O event for the user.
These views summarize stages, grouped by user. By default, rows are sorted by user and descending total stage latency.
The user_summary_by_stages and
x$user_summary_by_stages views
have these columns:
user
The client user name. Rows for which the
USER column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
event_name
The stage event name.
total
The total number of occurrences of the stage event for the user.
total_latency
The total wait time of timed occurrences of the stage event for the user.
avg_latency
The average wait time per timed occurrence of the stage event for the user.
These views summarize overall statement statistics, grouped by user. By default, rows are sorted by descending total latency.
The
user_summary_by_statement_latency
and
x$user_summary_by_statement_latency
views have these columns:
user
The client user name. Rows for which the
USER column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
total
The total number of statements for the user.
total_latency
The total wait time of timed statements for the user.
max_latency
The maximum single wait time of timed statements for the user.
lock_latency
The total time waiting for locks by timed statements for the user.
rows_sent
The total number of rows returned by statements for the user.
rows_examined
The total number of rows read from storage engines by statements for the user.
rows_affected
The total number of rows affected by statements for the user.
full_scans
The total number of full table scans by statements for the user.
These views summarize informaion about statements executed, grouped by user and statement type. By default, rows are sorted by user and descending total latency.
The
user_summary_by_statement_type
and
x$user_summary_by_statement_type
views have these columns:
user
The client user name. Rows for which the
USER column in the underlying
Performance Schema table is NULL are
assumed to be for background threads and are reported with
a host name of background.
statement
The final component of the statement event name.
total
The total number of occurrences of the statement event for the user.
total_latency
The total wait time of timed occurrences of the statement event for the user.
max_latency
The maximum single wait time of timed occurrences of the statement event for the user.
lock_latency
The total time waiting for locks by timed occurrences of the statement event for the user.
rows_sent
The total number of rows returned by occurrences of the statement event for the user.
rows_examined
The total number of rows read from storage engines by occurrences of the statement event for the user.
rows_affected
The total number of rows affected by occurrences of the statement event for the user.
full_scans
The total number of full table scans by occurrences of the statement event for the user.
This view provides the current
sys schema and MySQL server
versions.
The version view has these
columns:
sys_version
The sys schema version.
mysql_version
The MySQL server version.
These views summarize wait class average latencies, grouped by event class. By default, rows are sorted by descending average latency. Idle events are ignored.
An event class is determined by stripping from the event name
everything after the first three components. For example, the
class for wait/io/file/sql/slow_log is
wait/io/file.
The
wait_classes_global_by_avg_latency
and
x$wait_classes_global_by_avg_latency
views have these columns:
event_class
The event class.
total
The total number of occurrences of events in the class.
total_latency
The total wait time of timed occurrences of events in the class.
min_latency
The minimum single wait time of timed occurrences of events in the class.
avg_latency
The average wait time per timed occurrence of events in the class.
max_latency
The maximum single wait time of timed occurrences of events in the class.
These views summarize wait class total latencies, grouped by event class. By default, rows are sorted by descending total latency. Idle events are ignored.
An event class is determined by stripping from the event name
everything after the first three components. For example, the
class for wait/io/file/sql/slow_log is
wait/io/file.
The
wait_classes_global_by_latency
and
x$wait_classes_global_by_latency
views have these columns:
event_class
The event class.
total
The total number of occurrences of events in the class.
total_latency
The total wait time of timed occurrences of events in the class.
min_latency
The minimum single wait time of timed occurrences of events in the class.
avg_latency
The average wait time per timed occurrence of events in the class.
max_latency
The maximum single wait time of timed occurrences of events in the class.
These views summarize wait events, grouped by host and event. By default, rows are sorted by host and descending total latency. Idle events are ignored.
The waits_by_host_by_latency and
x$waits_by_host_by_latency views
have these columns:
host
The host from which the connection originated.
event
The event name.
total
The total number of occurrences of the event for the host.
total_latency
The total wait time of timed occurrences of the event for the host.
avg_latency
The average wait time per timed occurrence of the event for the host.
max_latency
The maximum single wait time of timed occurrences of the event for the host.
These views summarize wait events, grouped by user and event. By default, rows are sorted by user and descending total latency. Idle events are ignored.
The waits_by_user_by_latency and
x$waits_by_user_by_latency views
have these columns:
user
The user associated with the connection.
event
The event name.
total
The total number of occurrences of the event for the user.
total_latency
The total wait time of timed occurrences of the event for the user.
avg_latency
The average wait time per timed occurrence of the event for the user.
max_latency
The maximum single wait time of timed occurrences of the event for the user.
These views summarize wait events, grouped by event. By default, rows are sorted by descending total latency. Idle events are ignored.
The waits_global_by_latency and
x$waits_global_by_latency views
have these columns:
events
The event name.
total
The total number of occurrences of the event.
total_latency
The total wait time of timed occurrences of the event.
avg_latency
The average wait time per timed occurrence of the event.
max_latency
The maximum single wait time of timed occurrences of the event.
The following sections describe
sys schema stored procedures.
Given a schema name, this procedure creates a synonym schema
containing views that refer to all the tables and views in the
original schema. This can be used, for example, to create a
shorter name by which to refer to a schema with a long name
(such as info rather than
INFORMATION_SCHEMA).
in_db_name VARCHAR(64): The name of
the schema for which to create the synonym.
in_synonym VARCHAR(64): The name to
use for the synonym schema. This schema must not already
exist.
mysql>SHOW DATABASES;+--------------------+ | Database | +--------------------+ | information_schema | | mysql | | performance_schema | | sys | | world | +--------------------+ mysql>CALL sys.create_synonym_db('INFORMATION_SCHEMA', 'info');+---------------------------------------+ | summary | +---------------------------------------+ | Created 63 views in the info database | +---------------------------------------+ mysql>SHOW DATABASES;+--------------------+ | Database | +--------------------+ | information_schema | | info | | mysql | | performance_schema | | sys | | world | +--------------------+ mysql>SHOW FULL TABLES FROM info;+---------------------------------------+------------+ | Tables_in_info | Table_type | +---------------------------------------+------------+ | character_sets | VIEW | | collation_character_set_applicability | VIEW | | collations | VIEW | | column_privileges | VIEW | | columns | VIEW | ...
Creates a report of the current server status for diagnostic purposes.
This procedure disables binary logging during its execution by
manipulating the session value of the
sql_log_bin system variable.
That is a restricted operation, so the procedure requires
privileges sufficient to set restricted session variables. See
Section 5.1.9.1, “System Variable Privileges”.
Data collected for diagnostics()
includes this information:
Information from the metrics
view (see Section 27.4.3.21, “The metrics View”)
Information from other relevant sys
schema views, such as the one that detemines queries in
the 95th percentile
Information from the
ndbinfo schema, if the
MySQL server is part of NDB Cluster
Replication status (both master and slave)
Some of the sys schema views are calculated as initial (optional), overall, and delta values:
The initial view is the content of the view at the start
of the diagnostics()
procedure. This output is the same as the start values
used for the delta view. The initial view is included if
the diagnostics.include_raw
configuration option is ON.
The overall view is the content of the view at the end of
the diagnostics() procedure.
This output is the same as the end values used for the
delta view. The overall view is always included.
The delta view is the difference from the beginning to the
end of procedure execution. The minimum and maximum values
are the minimum and maximum values from the end view,
respectively. They do not necessarily reflect the minimum
and maximum values in the monitored period. Except for the
metrics view, the delta is
calculated only between the first and last outputs.
in_max_runtime INT UNSIGNED: The
maximum data collection time in seconds. Use
NULL to collect data for the default
of 60 seconds. Otherwise, use a value greater than 0.
in_interval INT UNSIGNED: The sleep
time between data collections in seconds. Use
NULL to sleep for the default of 30
seconds. Otherwise, use a value greater than 0.
in_auto_config ENUM('current', 'medium',
'full'): The Performance Schema configuration
to use. Permitted values are:
current: Use the current
instrument and consumer settings.
medium: Enable some instruments
and consumers.
full: Enable all instruments and
consumers.
The more instruments and consumers enabled, the more
impact on MySQL server performance. Be careful with
the medium setting and especially
the full setting, which has a large
performance impact.
Use of the medium or
full setting requires the
SUPER privilege.
If a setting other than current is
chosen, the current settings are restored at the end of
the procedure.
diagnostics() operation can be
modified using the following configuration options or their
corresponding user-defined variables (see
Section 27.4.2.1, “The sys_config Table”):
debug, @sys.debug
If this option is ON, produce
debugging output. The default is OFF.
diagnostics.allow_i_s_tables,
@sys.diagnostics.allow_i_s_tables
If this option is ON, the
diagnostics() procedure is
permitted to perform table scans on the
INFORMATION_SCHEMA.TABLES
table. This can be expensive if there are many tables.
The default is OFF.
diagnostics.include_raw,
@sys.diagnostics.include_raw
If this option is ON, the
diagnostics() procedure
output includes the raw output from querying the
metrics view. The default
is OFF.
statement_truncate_len,
@sys.statement_truncate_len
The maximum length of statements returned by the
format_statement()
function. Longer statements are truncated to this
length. The default is 64.
Create a diagnostics report that starts an iteration every 30 seconds and runs for at most 120 seconds using the current Performance Schema settings:
mysql> CALL sys.diagnostics(120, 30, 'current');
To capture the output from the
diagnostics() procedure in a file as it
runs, use the mysql client tee
and
filenamenotee commands (see
Section 4.5.1.2, “mysql Client Commands”):
mysql>tee diag.out;mysql>CALL sys.diagnostics(120, 30, 'current');mysql>notee;
Given an SQL statement as a string, executes it as a prepared statement. The prepared statement is deallocated after execution, so it is not subject to reuse. Thus, this procedure is useful primarily for executing dynamic statements on a one-time basis.
This procedure uses
sys_execute_prepared_stmt as the prepared
statement name. If that statement name exists when the
procedure is called, its previous content is destroyed.
execute_prepared_stmt()
operation can be modified using the following configuration
options or their corresponding user-defined variables (see
Section 27.4.2.1, “The sys_config Table”):
debug, @sys.debug
If this option is ON, produce
debugging output. The default is OFF.
Disables Performance Schema instrumentation for all background threads. Produces a result set indicating how many background threads were disabled. Already disabled threads do not count.
Disables Performance Schema consumers with names that contain the argument. Produces a result set indicating how many consumers were disabled. Already disabled consumers do not count.
consumer VARCHAR(128): The value used
to match consumer names, which are identified by using
%consumer% as an operand for a
LIKE pattern match.
A value of '' matches all consumers.
Disables Performance Schema instruments with names that contain the argument. Produces a result set indicating how many instruments were disabled. Already disabled instruments do not count.
in_pattern VARCHAR(128): The value
used to match instrument names, which are identified by
using %in_pattern% as an operand for
a LIKE pattern match.
A value of '' matches all
instruments.
Disable a specific instrument:
mysql> CALL sys.ps_setup_disable_instrument('wait/lock/metadata/sql/mdl');
+-----------------------+
| summary |
+-----------------------+
| Disabled 1 instrument |
+-----------------------+
Disable all mutex instruments:
mysql> CALL sys.ps_setup_disable_instrument('mutex');
+--------------------------+
| summary |
+--------------------------+
| Disabled 177 instruments |
+--------------------------+
Given a connection ID, disables Performance Schema instrumentation for the thread. Produces a result set indicating how many threads were disabled. Already disabled threads do not count.
in_connection_id BIGINT: The
connection ID. This is a connection ID as given in the
PROCESSLIST_ID column of the
Performance Schema threads
table or the Id column of
SHOW PROCESSLIST output.
Disable a specific connection by its connection ID:
mysql> CALL sys.ps_setup_disable_thread(225);
+-------------------+
| summary |
+-------------------+
| Disabled 1 thread |
+-------------------+
Disable the current connection:
mysql> CALL sys.ps_setup_disable_thread(CONNECTION_ID());
+-------------------+
| summary |
+-------------------+
| Disabled 1 thread |
+-------------------+
Enables Performance Schema instrumentation for all background threads. Produces a result set indicating how many background threads were enabled. Already enabled threads do not count.
Enables Performance Schema consumers with names that contain the argument. Produces a result set indicating how many consumers were enabled. Already enabled consumers do not count.
consumer VARCHAR(128): The value used
to match consumer names, which are identified by using
%consumer% as an operand for a
LIKE pattern match.
A value of '' matches all consumers.
Enables Performance Schema instruments with names that contain the argument. Produces a result set indicating how many instruments were enabled. Already enabled instruments do not count.
in_pattern VARCHAR(128): The value
used to match instrument names, which are identified by
using %in_pattern% as an operand for
a LIKE pattern match.
A value of '' matches all
instruments.
Enable a specific instrument:
mysql> CALL sys.ps_setup_enable_instrument('wait/lock/metadata/sql/mdl');
+----------------------+
| summary |
+----------------------+
| Enabled 1 instrument |
+----------------------+
Enable all mutex instruments:
mysql> CALL sys.ps_setup_enable_instrument('mutex');
+-------------------------+
| summary |
+-------------------------+
| Enabled 177 instruments |
+-------------------------+
Given a connection ID, enables Performance Schema instrumentation for the thread. Produces a result set indicating how many threads were enabled. Already enabled threads do not count.
in_connection_id BIGINT: The
connection ID. This is a connection ID as given in the
PROCESSLIST_ID column of the
Performance Schema threads
table or the Id column of
SHOW PROCESSLIST output.
Enable a specific connection by its connection ID:
mysql> CALL sys.ps_setup_enable_thread(225);
+------------------+
| summary |
+------------------+
| Enabled 1 thread |
+------------------+
Enable the current connection:
mysql> CALL sys.ps_setup_enable_thread(CONNECTION_ID());
+------------------+
| summary |
+------------------+
| Enabled 1 thread |
+------------------+
Reloads a Performance Schema configuration saved earlier
within the same session using
ps_setup_save(). For more
information, see the description of
ps_setup_save().
This procedure disables binary logging during its execution by
manipulating the session value of the
sql_log_bin system variable.
That is a restricted operation, so the procedure requires
privileges sufficient to set restricted session variables. See
Section 5.1.9.1, “System Variable Privileges”.
Resets the Performance Schema configuration to its default settings.
in_verbose BOOLEAN: Whether to
display information about each setup stage during
procedure execution. This includes the SQL statements
executed.
mysql> CALL sys.ps_setup_reset_to_default(TRUE)\G
*************************** 1. row ***************************
status: Resetting: setup_actors
DELETE
FROM performance_schema.setup_actors
WHERE NOT (HOST = '%' AND USER = '%' AND ROLE = '%')
*************************** 1. row ***************************
status: Resetting: setup_actors
INSERT IGNORE INTO performance_schema.setup_actors
VALUES ('%', '%', '%')
...
Saves the current Performance Schema configuration. This
enables you to alter the configuration temporarily for
debugging or other purposes, then restore it to the previous
state by invoking the
ps_setup_reload_saved()
procedure.
To prevent other simultaneous calls to save the configuration,
ps_setup_save() acquires an
advisory lock named sys.ps_setup_save by
calling the GET_LOCK()
function. ps_setup_save() takes
a timeout parameter to indicate how many seconds to wait if
the lock already exists (which indicates that some other
session has a saved configuration outstanding). If the timeout
expires without obtaining the lock,
ps_setup_save() fails.
It is intended you call
ps_setup_reload_saved() later
within the same session as
ps_setup_save() because the
configuration is saved in TEMPORARY tables.
ps_setup_save() drops the
temporary tables and releases the lock. If you end your
session without invoking
ps_setup_save(), the tables and
lock disappear automatically.
This procedure disables binary logging during its execution by
manipulating the session value of the
sql_log_bin system variable.
That is a restricted operation, so the procedure requires
privileges sufficient to set restricted session variables. See
Section 5.1.9.1, “System Variable Privileges”.
in_timeout INT: How many seconds to
wait to obtain the sys.ps_setup_save
lock. A negative timeout value means infinite timeout.
Displays all currently disabled Performance Schema configuration.
in_show_instruments BOOLEAN: Whether
to display disabled instruments. This might be a long
list.
in_show_threads BOOLEAN: Whether to
display disabled threads.
mysql> CALL sys.ps_setup_show_disabled(TRUE, TRUE); +----------------------------+ | performance_schema_enabled | +----------------------------+ | 1 | +----------------------------+ +---------------+ | enabled_users | +---------------+ | '%'@'%' | +---------------+ +-------------+----------------------+---------+-------+ | object_type | objects | enabled | timed | +-------------+----------------------+---------+-------+ | EVENT | mysql.% | NO | NO | | EVENT | performance_schema.% | NO | NO | | EVENT | information_schema.% | NO | NO | | FUNCTION | mysql.% | NO | NO | | FUNCTION | performance_schema.% | NO | NO | | FUNCTION | information_schema.% | NO | NO | | PROCEDURE | mysql.% | NO | NO | | PROCEDURE | performance_schema.% | NO | NO | | PROCEDURE | information_schema.% | NO | NO | | TABLE | mysql.% | NO | NO | | TABLE | performance_schema.% | NO | NO | | TABLE | information_schema.% | NO | NO | | TRIGGER | mysql.% | NO | NO | | TRIGGER | performance_schema.% | NO | NO | | TRIGGER | information_schema.% | NO | NO | +-------------+----------------------+---------+-------+ ...
Displays all currently disabled Performance Schema consumers.
mysql> CALL sys.ps_setup_show_disabled_consumers();
+----------------------------------+
| disabled_consumers |
+----------------------------------+
| events_stages_current |
| events_stages_history |
| events_stages_history_long |
| events_statements_history |
| events_statements_history_long |
| events_transactions_history |
| events_transactions_history_long |
| events_waits_current |
| events_waits_history |
| events_waits_history_long |
+----------------------------------+
Displays all currently disabled Performance Schema instruments. This might be a long list.
mysql> CALL sys.ps_setup_show_disabled_instruments()\G
*************************** 1. row ***************************
disabled_instruments: wait/synch/mutex/sql/TC_LOG_MMAP::LOCK_tc
timed: NO
*************************** 2. row ***************************
disabled_instruments: wait/synch/mutex/sql/THD::LOCK_query_plan
timed: NO
*************************** 3. row ***************************
disabled_instruments: wait/synch/mutex/sql/MYSQL_BIN_LOG::LOCK_commit
timed: NO
...
Displays all currently enabled Performance Schema configuration.
in_show_instruments BOOLEAN: Whether
to display enabled instruments. This might be a long
list.
in_show_threads BOOLEAN: Whether to
display enabled threads.
mysql> CALL sys.ps_setup_show_enabled(FALSE, FALSE);
+----------------------------+
| performance_schema_enabled |
+----------------------------+
| 1 |
+----------------------------+
1 row in set (0.01 sec)
+---------------+
| enabled_users |
+---------------+
| '%'@'%' |
+---------------+
1 row in set (0.01 sec)
+-------------+---------+---------+-------+
| object_type | objects | enabled | timed |
+-------------+---------+---------+-------+
| EVENT | %.% | YES | YES |
| FUNCTION | %.% | YES | YES |
| PROCEDURE | %.% | YES | YES |
| TABLE | %.% | YES | YES |
| TRIGGER | %.% | YES | YES |
+-------------+---------+---------+-------+
5 rows in set (0.02 sec)
+-----------------------------+
| enabled_consumers |
+-----------------------------+
| events_statements_current |
| events_statements_history |
| events_transactions_current |
| events_transactions_history |
| global_instrumentation |
| statements_digest |
| thread_instrumentation |
+-----------------------------+
Displays all currently enabled Performance Schema consumers.
mysql> CALL sys.ps_setup_show_enabled_consumers();
+-----------------------------+
| enabled_consumers |
+-----------------------------+
| events_statements_current |
| events_statements_history |
| events_transactions_current |
| events_transactions_history |
| global_instrumentation |
| statements_digest |
| thread_instrumentation |
+-----------------------------+
Displays all currently enabled Performance Schema instruments. This might be a long list.
mysql> CALL sys.ps_setup_show_enabled_instruments()\G
*************************** 1. row ***************************
enabled_instruments: wait/io/file/sql/map
timed: YES
*************************** 2. row ***************************
enabled_instruments: wait/io/file/sql/binlog
timed: YES
*************************** 3. row ***************************
enabled_instruments: wait/io/file/sql/binlog_cache
timed: YES
...
Displays a textual histogram graph of the average latency
values across all normalized statements tracked within the
Performance Schema
events_statements_summary_by_digest
table.
This procedure can be used to display a very high-level picture of the latency distribution of statements running within this MySQL instance.
The histogram output in statement units. For example,
* = 2 units in the histogram legend means
that each * character represents 2
statements.
mysql> CALL sys.ps_statement_avg_latency_histogram()\G
*************************** 1. row ***************************
Performance Schema Statement Digest Average Latency Histogram:
. = 1 unit
* = 2 units
# = 3 units
(0 - 66ms) 88 | #############################
(66 - 133ms) 14 | ..............
(133 - 199ms) 4 | ....
(199 - 265ms) 5 | **
(265 - 332ms) 1 | .
(332 - 398ms) 0 |
(398 - 464ms) 1 | .
(464 - 531ms) 0 |
(531 - 597ms) 0 |
(597 - 663ms) 0 |
(663 - 730ms) 0 |
(730 - 796ms) 0 |
(796 - 863ms) 0 |
(863 - 929ms) 0 |
(929 - 995ms) 0 |
(995 - 1062ms) 0 |
Total Statements: 114; Buckets: 16; Bucket Size: 66 ms;
Traces all Performance Schema instrumentation for a specific statement digest.
If you find a statement of interest within the Performance
Schema
events_statements_summary_by_digest
table, specify its DIGEST column MD5 value
to this procedure and indicate the polling duration and
interval. The result is a report of all statistics tracked
within Performance Schema for that digest for the interval.
The procedure also attempts to execute
EXPLAIN for the longest running
example of the digest during the interval. This attempt might
fail because the Performance Schema truncates long
SQL_TEXT values. Consequently,
EXPLAIN will fail due to parse
errors.
This procedure disables binary logging during its execution by
manipulating the session value of the
sql_log_bin system variable.
That is a restricted operation, so the procedure requires
privileges sufficient to set restricted session variables. See
Section 5.1.9.1, “System Variable Privileges”.
in_digest VARCHAR(32): The statement
digest identifier to analyze.
in_runtime INT: How long to run the
analysis in seconds.
in_interval DECIMAL(2,2): The
analysis interval in seconds (which can be fractional)
at which to try to take snapshots.
in_start_fresh BOOLEAN: Whether to
truncate the Performance Schema
events_statements_history_long
and
events_stages_history_long
tables before starting.
in_auto_enable BOOLEAN: Whether to
automatically enable required consumers.
mysql> CALL sys.ps_trace_statement_digest('891ec6860f98ba46d89dd20b0c03652c', 10, 0.1, TRUE, TRUE);
+--------------------+
| SUMMARY STATISTICS |
+--------------------+
| SUMMARY STATISTICS |
+--------------------+
1 row in set (9.11 sec)
+------------+-----------+-----------+-----------+---------------+------------+------------+
| executions | exec_time | lock_time | rows_sent | rows_examined | tmp_tables | full_scans |
+------------+-----------+-----------+-----------+---------------+------------+------------+
| 21 | 4.11 ms | 2.00 ms | 0 | 21 | 0 | 0 |
+------------+-----------+-----------+-----------+---------------+------------+------------+
1 row in set (9.11 sec)
+------------------------------------------+-------+-----------+
| event_name | count | latency |
+------------------------------------------+-------+-----------+
| stage/sql/statistics | 16 | 546.92 us |
| stage/sql/freeing items | 18 | 520.11 us |
| stage/sql/init | 51 | 466.80 us |
...
| stage/sql/cleaning up | 18 | 11.92 us |
| stage/sql/executing | 16 | 6.95 us |
+------------------------------------------+-------+-----------+
17 rows in set (9.12 sec)
+---------------------------+
| LONGEST RUNNING STATEMENT |
+---------------------------+
| LONGEST RUNNING STATEMENT |
+---------------------------+
1 row in set (9.16 sec)
+-----------+-----------+-----------+-----------+---------------+------------+-----------+
| thread_id | exec_time | lock_time | rows_sent | rows_examined | tmp_tables | full_scan |
+-----------+-----------+-----------+-----------+---------------+------------+-----------+
| 166646 | 618.43 us | 1.00 ms | 0 | 1 | 0 | 0 |
+-----------+-----------+-----------+-----------+---------------+------------+-----------+
1 row in set (9.16 sec)
# Truncated for clarity...
+-----------------------------------------------------------------+
| sql_text |
+-----------------------------------------------------------------+
| select hibeventhe0_.id as id1382_, hibeventhe0_.createdTime ... |
+-----------------------------------------------------------------+
1 row in set (9.17 sec)
+------------------------------------------+-----------+
| event_name | latency |
+------------------------------------------+-----------+
| stage/sql/init | 8.61 us |
| stage/sql/init | 331.07 ns |
...
| stage/sql/freeing items | 30.46 us |
| stage/sql/cleaning up | 662.13 ns |
+------------------------------------------+-----------+
18 rows in set (9.23 sec)
+----+-------------+--------------+-------+---------------+-----------+---------+-------------+------+-------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+--------------+-------+---------------+-----------+---------+-------------+------+-------+
| 1 | SIMPLE | hibeventhe0_ | const | fixedTime | fixedTime | 775 | const,const | 1 | NULL |
+----+-------------+--------------+-------+---------------+-----------+---------+-------------+------+-------+
1 row in set (9.27 sec)
Query OK, 0 rows affected (9.28 sec)
Dumps all Performance Schema data for an instrumented thread
to a .dot formatted graph file (for the
DOT graph description language). Each result set returned from
the procedure should be used for a complete graph.
This procedure disables binary logging during its execution by
manipulating the session value of the
sql_log_bin system variable.
That is a restricted operation, so the procedure requires
privileges sufficient to set restricted session variables. See
Section 5.1.9.1, “System Variable Privileges”.
in_thread_id INT: The thread to
trace.
in_outfile VARCHAR(255): The name to
use for the .dot output file.
in_max_runtime DECIMAL(20,2): The
maximum number of seconds (which can be fractional) to
collect data. Use NULL to collect
data for the default of 60 seconds.
in_interval DECIMAL(20,2): The number
of seconds (which can be fractional) to sleep between
data collections. Use NULL to sleep
for the default of 1 second.
in_start_fresh BOOLEAN: Whether to
reset all Performance Schema data before tracing.
in_auto_setup BOOLEAN: Whether to
disable all other threads and enable all instruments and
consumers. This also resets the settings at the end of
the run.
in_debug BOOLEAN: Whether to include
file:lineno information in the graph.
mysql> CALL sys.ps_trace_thread(25, CONCAT('/tmp/stack-', REPLACE(NOW(), ' ', '-'), '.dot'), NULL, NULL, TRUE, TRUE, TRUE);
+-------------------+
| summary |
+-------------------+
| Disabled 1 thread |
+-------------------+
1 row in set (0.00 sec)
+---------------------------------------------+
| Info |
+---------------------------------------------+
| Data collection starting for THREAD_ID = 25 |
+---------------------------------------------+
1 row in set (0.03 sec)
+-----------------------------------------------------------+
| Info |
+-----------------------------------------------------------+
| Stack trace written to /tmp/stack-2014-02-16-21:18:41.dot |
+-----------------------------------------------------------+
1 row in set (60.07 sec)
+-------------------------------------------------------------------+
| Convert to PDF |
+-------------------------------------------------------------------+
| dot -Tpdf -o /tmp/stack_25.pdf /tmp/stack-2014-02-16-21:18:41.dot |
+-------------------------------------------------------------------+
1 row in set (60.07 sec)
+-------------------------------------------------------------------+
| Convert to PNG |
+-------------------------------------------------------------------+
| dot -Tpng -o /tmp/stack_25.png /tmp/stack-2014-02-16-21:18:41.dot |
+-------------------------------------------------------------------+
1 row in set (60.07 sec)
+------------------+
| summary |
+------------------+
| Enabled 1 thread |
+------------------+
1 row in set (60.32 sec)
Truncates all Performance Schema summary tables, resetting all aggregated instrumentation as a snapshot. Produces a result set indicating how many tables were truncated.
in_verbose BOOLEAN: Whether to
display each TRUNCATE
TABLE statement before executing it.
Creates a report of the statements running on the server. The views are calculated based on the overall and/or delta activity.
This procedure disables binary logging during its execution by
manipulating the session value of the
sql_log_bin system variable.
That is a restricted operation, so the procedure requires
privileges sufficient to set restricted session variables. See
Section 5.1.9.1, “System Variable Privileges”.
in_action ENUM('snapshot', 'overall', 'delta',
'create_tmp', 'create_table', 'save',
'cleanup'): The action to take. These values
are permitted:
snapshot: Store a snapshot. The
default is to make a snapshot of the current content
of the Performance Schema
events_statements_summary_by_digest
table. By setting in_table, this
can be overwritten to copy the content of the
specified table. The snapshot is stored in the
sys schema
tmp_digests temporary table.
overall: Generate an analysis
based on the content of the table specified by
in_table. For the overall
analysis, in_table can be
NOW() to use a fresh
snapshot. This overwrites an existing snapshot. Use
NULL for
in_table to use the existing
snapshot. If in_table is
NULL and no snapshot exists, a
new snapshot is created. The
in_views parameter and the
statement_performance_analyzer.limit
configuration option affect the operation of this
procedure.
delta: Generate a delta analysis.
The delta is calculated between the reference table
specified by in_table and the
snapshot, which must exist. This action uses the
sys schema
tmp_digests_delta temporary
table. The in_views parameter and
the
statement_performance_analyzer.limit
configuration option affect the operation of this
procedure.
create_table: Create a regular
table suitable for storing the snapshot for later
use (for example, for calculating deltas).
create_tmp: Create a temporary
table suitable for storing the snapshot for later
use (for example, for calculating deltas).
save: Save the snapshot in the
table specified by in_table. The
table must exist and have the correct structure. If
no snapshot exists, a new snapshot is created.
cleanup: Remove the temporary
tables used for the snapshot and delta.
in_table VARCHAR(129): The table
parameter used for some of the actions specified by the
in_action parameter. Use the format
db_name.tbl_name or
tbl_name without using any
backtick (`) identifier-quoting
characters. Periods (.) are not
supported in database and table names.
The meaning of the in_table value for
each in_action value is detailed in
the individual in_action value
descriptions.
in_views SET
('with_runtimes_in_95th_percentile', 'analysis',
'with_errors_or_warnings', 'with_full_table_scans',
'with_sorting', 'with_temp_tables', 'custom'):
Which views to include. This parameter is a
SET value, so it can contain multiple
view names, separated by commas. The default is to
include all views except custom. The
following values are permitted:
with_runtimes_in_95th_percentile:
Use the
statements_with_runtimes_in_95th_percentile
view.
analysis: Use the
statement_analysis
view.
with_errors_or_warnings: Use the
statements_with_errors_or_warnings
view.
with_full_table_scans: Use the
statements_with_full_table_scans
view.
with_sorting: Use the
statements_with_sorting
view.
with_temp_tables: Use the
statements_with_temp_tables
view.
custom: Use a custom view. This
view must be specified using the
statement_performance_analyzer.view
configuration option to name a query or an existing
view.
statement_performance_analyzer()
operation can be modified using the following configuration
options or their corresponding user-defined variables (see
Section 27.4.2.1, “The sys_config Table”):
debug, @sys.debug
If this option is ON, produce
debugging output. The default is OFF.
statement_performance_analyzer.limit,
@sys.statement_performance_analyzer.limit
The maximum number of rows to return for views that have no built-in limit. The default is 100.
statement_performance_analyzer.view,
@sys.statement_performance_analyzer.view
The custom query or view to be used. If the option value
contains a space, it is interpreted as a query.
Otherwise, it must be the name of an existing view that
queries the Performance Schema
events_statements_summary_by_digest
table. There cannot be any LIMIT
clause in the query or view definition if the
statement_performance_analyzer.limit
configuration option is greater than 0. If specifying a
view, use the same format as for the
in_table parameter. The default is
NULL (no custom view defined).
To create a report with the queries in the 95th percentile
since the last truncation of
events_statements_summary_by_digest
and with a one-minute delta period:
Create a temporary table to store the initial snapshot.
Create the initial snapshot.
Save the initial snapshot in the temporary table.
Wait one minute.
Create a new snapshot.
Perform analysis based on the new snapshot.
Perform analysis based on the delta between the initial and new snapshots.
mysql>CALL sys.statement_performance_analyzer('create_tmp', 'mydb.tmp_digests_ini', NULL);Query OK, 0 rows affected (0.08 sec) mysql>CALL sys.statement_performance_analyzer('snapshot', NULL, NULL);Query OK, 0 rows affected (0.02 sec) mysql>CALL sys.statement_performance_analyzer('save', 'mydb.tmp_digests_ini', NULL);Query OK, 0 rows affected (0.00 sec) mysql>DO SLEEP(60);Query OK, 0 rows affected (1 min 0.00 sec) mysql>CALL sys.statement_performance_analyzer('snapshot', NULL, NULL);Query OK, 0 rows affected (0.02 sec) mysql>CALL sys.statement_performance_analyzer('overall', NULL, 'with_runtimes_in_95th_percentile');+-----------------------------------------+ | Next Output | +-----------------------------------------+ | Queries with Runtime in 95th Percentile | +-----------------------------------------+ 1 row in set (0.05 sec) ... mysql>CALL sys.statement_performance_analyzer('delta', 'mydb.tmp_digests_ini', 'with_runtimes_in_95th_percentile');+-----------------------------------------+ | Next Output | +-----------------------------------------+ | Queries with Runtime in 95th Percentile | +-----------------------------------------+ 1 row in set (0.03 sec) ...
Create an overall report of the 95th percentile queries and the top 10 queries with full table scans:
mysql>CALL sys.statement_performance_analyzer('snapshot', NULL, NULL);Query OK, 0 rows affected (0.01 sec) mysql>SET @sys.statement_performance_analyzer.limit = 10;Query OK, 0 rows affected (0.00 sec) mysql>CALL sys.statement_performance_analyzer('overall', NULL, 'with_runtimes_in_95th_percentile,with_full_table_scans');+-----------------------------------------+ | Next Output | +-----------------------------------------+ | Queries with Runtime in 95th Percentile | +-----------------------------------------+ 1 row in set (0.01 sec) ... +-------------------------------------+ | Next Output | +-------------------------------------+ | Top 10 Queries with Full Table Scan | +-------------------------------------+ 1 row in set (0.09 sec) ...
Use a custom view showing the top 10 queries sorted by total execution time, refreshing the view every minute using the watch command in Linux:
mysql>CREATE OR REPLACE VIEW mydb.my_statements ASSELECT sys.format_statement(DIGEST_TEXT) AS query,SCHEMA_NAME AS db,COUNT_STAR AS exec_count,sys.format_time(SUM_TIMER_WAIT) AS total_latency,sys.format_time(AVG_TIMER_WAIT) AS avg_latency,ROUND(IFNULL(SUM_ROWS_SENT / NULLIF(COUNT_STAR, 0), 0)) AS rows_sent_avg,ROUND(IFNULL(SUM_ROWS_EXAMINED / NULLIF(COUNT_STAR, 0), 0)) AS rows_examined_avg,ROUND(IFNULL(SUM_ROWS_AFFECTED / NULLIF(COUNT_STAR, 0), 0)) AS rows_affected_avg,DIGEST AS digestFROM performance_schema.events_statements_summary_by_digestORDER BY SUM_TIMER_WAIT DESC;Query OK, 0 rows affected (0.10 sec) mysql>CALL sys.statement_performance_analyzer('create_table', 'mydb.digests_prev', NULL);Query OK, 0 rows affected (0.10 sec) shell>watch -n 60 "mysql sys --table -e \">SET @sys.statement_performance_analyzer.view = 'mydb.my_statements';>SET @sys.statement_performance_analyzer.limit = 10;>CALL statement_performance_analyzer('snapshot', NULL, NULL);>CALL statement_performance_analyzer('delta', 'mydb.digests_prev', 'custom');>CALL statement_performance_analyzer('save', 'mydb.digests_prev', NULL);>\""Every 60.0s: mysql sys --table -e " ... Mon Dec 22 10:58:51 2014 +----------------------------------+ | Next Output | +----------------------------------+ | Top 10 Queries Using Custom View | +----------------------------------+ +-------------------+-------+------------+---------------+-------------+---------------+-------------------+-------------------+----------------------------------+ | query | db | exec_count | total_latency | avg_latency | rows_sent_avg | rows_examined_avg | rows_affected_avg | digest | +-------------------+-------+------------+---------------+-------------+---------------+-------------------+-------------------+----------------------------------+ ...
Tests whether a given table exists as a regular table, a
TEMPORARY table, or a view. The procedure
returns the table type in an OUT parameter.
If both a temporary and a permanent table exist with the given
name, TEMPORARY is returned.
in_db VARCHAR(64): The name of the
database in which to check for table existance.
in_table VARCHAR(64): The name of the
table to check the existance of.
out_exists ENUM('', 'BASE TABLE', 'VIEW',
'TEMPORARY'): The return value. This is an
OUT parameter, so it must be a
variable into which the table type can be stored. When
the procedure returns, the variable has one of the
following values to indicate whether the table exists:
'': The table name does not exist
as a base table, TEMPORARY table,
or view.
BASE TABLE: The table name exists
as a base (permanent) table.
VIEW: The table name exists as a
view.
TEMPORARY: The table name exists
as a TEMPORARY table.
mysql>CREATE DATABASE db1;Query OK, 1 row affected (0.01 sec) mysql>USE db1;Database changed mysql>CREATE TABLE t1 (id INT PRIMARY KEY);Query OK, 0 rows affected (0.03 sec) mysql>CREATE TABLE t2 (id INT PRIMARY KEY);Query OK, 0 rows affected (0.20 sec) mysql>CREATE view v_t1 AS SELECT * FROM t1;Query OK, 0 rows affected (0.02 sec) mysql>CREATE TEMPORARY TABLE t1 (id INT PRIMARY KEY);Query OK, 0 rows affected (0.00 sec) mysql>CALL sys.table_exists('db1', 't1', @exists); SELECT @exists;Query OK, 0 rows affected (0.01 sec) +-----------+ | @exists | +-----------+ | TEMPORARY | +-----------+ 1 row in set (0.00 sec) mysql>CALL sys.table_exists('db1', 't2', @exists); SELECT @exists;Query OK, 0 rows affected (0.02 sec) +------------+ | @exists | +------------+ | BASE TABLE | +------------+ 1 row in set (0.00 sec) mysql>CALL sys.table_exists('db1', 'v_t1', @exists); SELECT @exists;Query OK, 0 rows affected (0.02 sec) +---------+ | @exists | +---------+ | VIEW | +---------+ 1 row in set (0.00 sec) mysql>CALL sys.table_exists('db1', 't3', @exists); SELECT @exists;Query OK, 0 rows affected (0.00 sec) +---------+ | @exists | +---------+ | | +---------+ 1 row in set (0.00 sec)
The following sections describe
sys schema stored functions.
Given a file path name, returns the path component that
represents the schema name. This function assumes that the
file name lies within the schema directory. For this reason,
it will not work with partitions or tables defined using their
own DATA_DIRECTORY table option.
This function is useful when extracting file I/O information from the Performance Schema that includes file path names. It provides a convenient way to display schema names, which can be more easily understood than full path names, and can be used in joins against object schema names.
mysql> SELECT sys.extract_schema_from_file_name('/usr/local/mysql/data/world/City.ibd');
+---------------------------------------------------------------------------+
| sys.extract_schema_from_file_name('/usr/local/mysql/data/world/City.ibd') |
+---------------------------------------------------------------------------+
| world |
+---------------------------------------------------------------------------+
Given a file path name, returns the path component that represents the table name.
This function is useful when extracting file I/O information from the Performance Schema that includes file path names. It provides a convenient way to display table names, which can be more easily understood than full path names, and can be used in joins against object table names.
mysql> SELECT sys.extract_table_from_file_name('/usr/local/mysql/data/world/City.ibd');
+--------------------------------------------------------------------------+
| sys.extract_table_from_file_name('/usr/local/mysql/data/world/City.ibd') |
+--------------------------------------------------------------------------+
| City |
+--------------------------------------------------------------------------+
Given a value in bytes, converts it to human-readable format
and returns a string consisting of a value and a units
indicator. Depending on the size of the value, the units part
is bytes, KiB
(kibibytes), MiB (mebibytes),
GiB (gibibytes), TiB
(tebibytes), or PiB (pebibytes).
mysql> SELECT sys.format_bytes(512), sys.format_bytes(18446644073709551615);
+-----------------------+----------------------------------------+
| sys.format_bytes(512) | sys.format_bytes(18446644073709551615) |
+-----------------------+----------------------------------------+
| 512 bytes | 16383.91 PiB |
+-----------------------+----------------------------------------+
Given a path name, returns the modified path name after replacing subpaths that match the values of the following system variables, in order:
datadir tmpdir slave_load_tmpdir innodb_data_home_dir innodb_log_group_home_dir innodb_undo_directory basedir
A value that matches the value of system variable
sysvar is replaced with the string
@@GLOBAL..
sysvar
mysql> SELECT sys.format_path('/usr/local/mysql/data/world/City.ibd');
+---------------------------------------------------------+
| sys.format_path('/usr/local/mysql/data/world/City.ibd') |
+---------------------------------------------------------+
| /usr/local/mysql/data/world/City.ibd |
+---------------------------------------------------------+
Given a string (normally representing an SQL statement),
reduces it to the length given by the
statement_truncate_len configuration
option, and returns the result. No truncation occurs if the
string is shorter than
statement_truncate_len. Otherwise, the
middle part of the string is replaced by an ellipsis
(...).
This function is useful for formatting possibly lengthy statements retrieved from Performance Schema tables to a known fixed maximum length.
format_statement() operation
can be modified using the following configuration options or
their corresponding user-defined variables (see
Section 27.4.2.1, “The sys_config Table”):
statement_truncate_len,
@sys.statement_truncate_len
The maximum length of statements returned by the
format_statement()
function. Longer statements are truncated to this
length. The default is 64.
By default, format_statement()
truncates statements to be no more than 64 characters.
Setting @sys.statement_truncate_len
changes the truncation length for the current session:
mysql>SET @stmt = 'SELECT variable, value, set_time, set_by FROM sys_config';mysql>SELECT sys.format_statement(@stmt);+----------------------------------------------------------+ | sys.format_statement(@stmt) | +----------------------------------------------------------+ | SELECT variable, value, set_time, set_by FROM sys_config | +----------------------------------------------------------+ mysql>SET @sys.statement_truncate_len = 32;mysql>SELECT sys.format_statement(@stmt);+-----------------------------------+ | sys.format_statement(@stmt) | +-----------------------------------+ | SELECT variabl ... ROM sys_config | +-----------------------------------+
Given a Performance Schema latency or wait time in
picoseconds, converts it to human-readable format and returns
a string consisting of a value and a units indicator.
Depending on the size of the value, the units part is
ns (nanoseconds), us
(microseconds), ms (milliseconds),
s (seconds), m
(minutes), h (hours), d
(days), or w (weeks).
mysql> SELECT sys.format_time(3501), sys.format_time(188732396662000);
+-----------------------+----------------------------------+
| sys.format_time(3501) | sys.format_time(188732396662000) |
+-----------------------+----------------------------------+
| 3.50 ns | 3.15 m |
+-----------------------+----------------------------------+
Adds a value to a comma-separated list of values and returns the result.
This function and list_drop()
can be useful for manipulating the value of system variables
such as sql_mode and
optimizer_switch that take a
comma-separated list of values.
mysql>SELECT @@sql_mode;+----------------------------------------+ | @@sql_mode | +----------------------------------------+ | ONLY_FULL_GROUP_BY,STRICT_TRANS_TABLES | +----------------------------------------+ mysql>SET @@sql_mode = sys.list_add(@@sql_mode, 'NO_ENGINE_SUBSTITUTION');mysql>SELECT @@sql_mode;+---------------------------------------------------------------+ | @@sql_mode | +---------------------------------------------------------------+ | ONLY_FULL_GROUP_BY,STRICT_TRANS_TABLES,NO_ENGINE_SUBSTITUTION | +---------------------------------------------------------------+ mysql>SET @@sql_mode = sys.list_drop(@@sql_mode, 'ONLY_FULL_GROUP_BY');mysql>SELECT @@sql_mode;+--------------------------------------------+ | @@sql_mode | +--------------------------------------------+ | STRICT_TRANS_TABLES,NO_ENGINE_SUBSTITUTION | +--------------------------------------------+
Removes a value from a comma-separated list of values and
returns the result. For more information, see the description
of list_add()
in_list TEXT: The list to be
modified.
in_drop_value TEXT: The value to drop
from the list.
Returns YES or NO to
indicate whether Performance Schema instrumentation for a
given account is enabled.
in_host VARCHAR(60): The host name of
the account to check.
in_user VARCHAR(32): The user name of
the account to check.
Returns YES or NO to
indicate whether a given Performance Schema consumer is
enabled, or NULL if the argument is not a
valid consumer name.
This function accounts for the consumer hierarchy, so a consumer is not considered enabled unless all consumers on which depends are also enabled. For information about the consumer hierarchy, see Section 26.4.7, “Pre-Filtering by Consumer”.
mysql> SELECT sys.ps_is_consumer_enabled('thread_instrumentation');
+------------------------------------------------------+
| sys.ps_is_consumer_enabled('thread_instrumentation') |
+------------------------------------------------------+
| YES |
+------------------------------------------------------+
Returns YES or NO to
indicate whether a given Performance Schema instrument is
enabled by default.
mysql>SELECT sys.ps_is_instrument_default_enabled('memory/innodb/row_log_buf');+-------------------------------------------------------------------+ | sys.ps_is_instrument_default_enabled('memory/innodb/row_log_buf') | +-------------------------------------------------------------------+ | NO | +-------------------------------------------------------------------+ mysql>SELECT sys.ps_is_instrument_default_enabled('statement/sql/alter_user');+------------------------------------------------------------------+ | sys.ps_is_instrument_default_enabled('statement/sql/alter_user') | +------------------------------------------------------------------+ | YES | +------------------------------------------------------------------+
Returns YES or NO to
indicate whether a given Performance Schema instrument is
timed by default.
mysql>SELECT sys.ps_is_instrument_default_timed('memory/innodb/row_log_buf');+-----------------------------------------------------------------+ | sys.ps_is_instrument_default_timed('memory/innodb/row_log_buf') | +-----------------------------------------------------------------+ | NO | +-----------------------------------------------------------------+ mysql>SELECT sys.ps_is_instrument_default_timed('statement/sql/alter_user');+----------------------------------------------------------------+ | sys.ps_is_instrument_default_timed('statement/sql/alter_user') | +----------------------------------------------------------------+ | YES | +----------------------------------------------------------------+
Returns YES or NO to
indicate whether Performance Schema instrumentation for a
given connection ID is enabled, UNKNOWN if
the ID is unknown, or NULL if the ID is
NULL.
in_connection_id BIGINT UNSIGNED: The
connection ID. This is a connection ID as given in the
PROCESSLIST_ID column of the
Performance Schema threads
table or the Id column of
SHOW PROCESSLIST output.
mysql>SELECT sys.ps_is_thread_instrumented(43);+-----------------------------------+ | sys.ps_is_thread_instrumented(43) | +-----------------------------------+ | UNKNOWN | +-----------------------------------+ mysql>SELECT sys.ps_is_thread_instrumented(CONNECTION_ID());+------------------------------------------------+ | sys.ps_is_thread_instrumented(CONNECTION_ID()) | +------------------------------------------------+ | YES | +------------------------------------------------+
Given a Performance Schema thread ID, returns the
user_name@host_name
in_thread_id BIGINT UNSIGNED: The
thread ID for which to return the account. The value
should match the THREAD_ID column
from some Performance Schema
threads table row.
mysql> SELECT sys.ps_thread_account(sys.ps_thread_id(CONNECTION_ID()));
+----------------------------------------------------------+
| sys.ps_thread_account(sys.ps_thread_id(CONNECTION_ID())) |
+----------------------------------------------------------+
| root@localhost |
+----------------------------------------------------------+
Returns the Performance Schema thread ID for a given
connection ID, or the thread ID for the current connection if
the connection ID is NULL.
in_connection_id BIGINT UNSIGNED: The
ID of the connection for which to return the thread ID.
This is a connection ID as given in the
PROCESSLIST_ID column of the
Performance Schema threads
table or the Id column of
SHOW PROCESSLIST output.
Returns a JSON formatted stack of all statements, stages, and events within the Performance Schema for a given thread ID.
in_thread_id BIGINT: The ID of the
thread to trace. The value should match the
THREAD_ID column from some
Performance Schema threads
table row.
in_verbose BOOLEAN: Whether to
include file:lineno information in
the events.
mysql> SELECT sys.ps_thread_stack(37, FALSE) AS thread_stack\G
*************************** 1. row ***************************
thread_stack: {"rankdir": "LR","nodesep": "0.10",
"stack_created": "2014-02-19 13:39:03", "mysql_version": "8.0.2-dmr-debug-log",
"mysql_user": "root@localhost","events": [{"nesting_event_id": "0",
"event_id": "10", "timer_wait": 256.35, "event_info": "sql/select",
"wait_info": "select @@version_comment limit 1\nerrors: 0\nwarnings: 0\nlock time:
...
Returns a JSON object containing information about a given
thread. The information includes the current transaction, and
the statements it has already executed, derived from the
Performance Schema
events_transactions_current and
events_statements_history tables.
(The consumers for those tables must be enabled to obtain full
data in the JSON object.)
If the output exceeds the truncation length (65535 by default), a JSON error object is returned, such as:
{ "error": "Trx info truncated: Row 6 was cut by GROUP_CONCAT()" }
Similar error objects are returned for other warnings and exceptions raised during function execution.
in_thread_id BIGINT UNSIGNED: The
thread ID for which to return transaction information.
The value should match the THREAD_ID
column from some Performance Schema
threads table row.
ps_thread_trx_info() operation
can be modified using the following configuration options or
their corresponding user-defined variables (see
Section 27.4.2.1, “The sys_config Table”):
ps_thread_trx_info.max_length,
@sys.ps_thread_trx_info.max_length
The maximum length of the output. The default is 65535.
mysql> SELECT sys.ps_thread_trx_info(48)\G
*************************** 1. row ***************************
sys.ps_thread_trx_info(48): [
{
"time": "790.70 us",
"state": "COMMITTED",
"mode": "READ WRITE",
"autocommitted": "NO",
"gtid": "AUTOMATIC",
"isolation": "REPEATABLE READ",
"statements_executed": [
{
"sql_text": "INSERT INTO info VALUES (1, \'foo\')",
"time": "471.02 us",
"schema": "trx",
"rows_examined": 0,
"rows_affected": 1,
"rows_sent": 0,
"tmp_tables": 0,
"tmp_disk_tables": 0,
"sort_rows": 0,
"sort_merge_passes": 0
},
{
"sql_text": "COMMIT",
"time": "254.42 us",
"schema": "trx",
"rows_examined": 0,
"rows_affected": 0,
"rows_sent": 0,
"tmp_tables": 0,
"tmp_disk_tables": 0,
"sort_rows": 0,
"sort_merge_passes": 0
}
]
},
{
"time": "426.20 us",
"state": "COMMITTED",
"mode": "READ WRITE",
"autocommitted": "NO",
"gtid": "AUTOMATIC",
"isolation": "REPEATABLE READ",
"statements_executed": [
{
"sql_text": "INSERT INTO info VALUES (2, \'bar\')",
"time": "107.33 us",
"schema": "trx",
"rows_examined": 0,
"rows_affected": 1,
"rows_sent": 0,
"tmp_tables": 0,
"tmp_disk_tables": 0,
"sort_rows": 0,
"sort_merge_passes": 0
},
{
"sql_text": "COMMIT",
"time": "213.23 us",
"schema": "trx",
"rows_examined": 0,
"rows_affected": 0,
"rows_sent": 0,
"tmp_tables": 0,
"tmp_disk_tables": 0,
"sort_rows": 0,
"sort_merge_passes": 0
}
]
}
]
Given a string argument, this function produces a quoted
identifier suitable for inclusion in SQL statements. This is
useful when a value to be used as an identifier is a reserved
word or contains backtick (`) characters.
mysql>SELECT sys.quote_identifier('plain');+-------------------------------+ | sys.quote_identifier('plain') | +-------------------------------+ | `plain` | +-------------------------------+ mysql>SELECT sys.quote_identifier('trick`ier');+-----------------------------------+ | sys.quote_identifier('trick`ier') | +-----------------------------------+ | `trick``ier` | +-----------------------------------+ mysql>SELECT sys.quote_identifier('integer');+---------------------------------+ | sys.quote_identifier('integer') | +---------------------------------+ | `integer` | +---------------------------------+
Given a configuration option name, returns the option value
from the sys_config table, or
the provided default value (which may be
NULL) if the option does not exist in the
table.
If sys_get_config() returns the
default value and that value is NULL, it is
expected that the caller is able to handle
NULL for the given configuration option.
By convention, routines that call
sys_get_config() first check
whether the corresponding user-defined variable exists and is
non-NULL. If so, the routine uses the
variable value without reading the
sys_config table. If the
variable does not exist or is NULL, the
routine reads the option value from the table and sets the
user-defined variable to that value. For more information
about the relationship between configuration options and their
corresponding user-defined variables, see
Section 27.4.2.1, “The sys_config Table”.
If you want to check whether the configuration option has
already been set and, if not, use the return value of
sys_get_config(), you can use
IFNULL(...) (see example later). However,
this should not be done inside a loop (for example, for each
row in a result set) because for repeated calls where the
assignment is needed only in the first iteration, using
IFNULL(...) is expected to be significantly
slower than using an IF (...) THEN ... END
IF; block (see example later).
in_variable_name VARCHAR(128): The
name of the configuration option for which to return the
value.
in_default_value VARCHAR(128): The
default value to return if the configuration option is
not found in the sys_config table.
Get a configuration value from the
sys_config table, falling back
to 128 as the default if the option is not present in the
table:
mysql> SELECT sys.sys_get_config('statement_truncate_len', 128) AS Value;
+-------+
| Value |
+-------+
| 64 |
+-------+
One-liner example: Check whether the option is already set;
if not, assign the IFNULL(...) result
(using the value from the
sys_config table):
mysql>SET @sys.statement_truncate_len =IFNULL(@sys.statement_truncate_len,sys.sys_get_config('statement_truncate_len', 64));
IF (...) THEN ... END IF; block example:
Check whether the option is already set; if not, assign the
value from the sys_config
table:
IF (@sys.statement_truncate_len IS NULL) THEN
SET @sys.statement_truncate_len = sys.sys_get_config('statement_truncate_len', 64);
END IF;