LP #1186375: long_query_time not respected in Solaris 11

Description

**Reported in Launchpad by Fernando Ipar last update 03-08-2013 04:17:33

I am aware of 1016991, but this situation seems different.

OS/MySQL versions:

SunOS <...supressed...> 5.11 <...supressed...> i86pc i386 i86pc Solaris
Server version: 5.5.27-log Source distribution

my.cnf:

[client]
port = 3306
socket = /tmp/mysql.sock

[mysqld]
user = mysql
port = 3306
socket = /tmp/mysql.sock
bind-address = 0.0.0.0

skip-external-locking
log_warnings
skip_name_resolv

basedir = /mysql
datadir = /databases

server-id = 1414
auto_increment_increment = 2
auto_increment_offset = 1

character-set-server = utf8
collation-server = utf8_unicode_ci

  1. InnoDB settings
    innodb_data_home_dir = /databases
    innodb_log_group_home_dir = /databases

innodb_data_file_path = ibdata1:100M:autoextend

innodb_buffer_pool_size = 70000M
innodb_buffer_pool_instances = 4

innodb_log_file_size = 400M
innodb_additional_mem_pool_size = 20M
innodb_log_buffer_size = 8M
innodb_flush_log_at_trx_commit = 2
innodb_lock_wait_timeout = 20
innodb_file_per_table
innodb_doublewrite = 0
innodb_io_capacity = 20000
innodb_read_io_threads = 48
innodb_write_io_threads = 48

  1. All other settings
    table_cache = 256
    table_open_cache = 1024
    table_definition_cache = 1024
    thread_cache_size = 250

query_cache_type = OFF

query_cache_size = 0
query_cache_limit = 4096
query_cache_min_res_unit = 512

thread_concurrency = 48
tmpdir = /tmp
tmp_table_size = 2048M
max_heap_table_size = 256M
max_connections = 1500
max_allowed_packet = 32M
connect_timeout = 10
net_read_timeout = 10
net_write_timeout = 10
net_buffer_length = 16384
read_buffer_size = 16M
thread_stack = 384K
transaction-isolation = READ-UNCOMMITTED
event_scheduler = OFF
group_concat_max_len = 1024
slave_type_conversions =

  1. Slow query log settings
    slow_query_log = OFF
    slow_query_log_file = /var/log/mysql/slowquery.log
    slow_query_log_timestamp_always = ON
    slow_query_log_timestamp_precision = microsecond
    slow_query_log_use_global_control = all
    long_query_time = 0
    log_slow_filter =
    log_slow_verbosity = microtime,query_plan,innodb
    log-queries-not-using-indexes = OFF

  1. Replication settings (master to slave)
    max_binlog_size = 1000M
    max-relay-log-size = 1000M
    log-slave-updates
    log-bin = /var/log/mysql/bin.log
    log-error = /var/log/mysql/error.log
    relay-log = /var/log/mysql/relay.log
    expire_logs_days = 3
    binlog_format = "ROW"

[mysqldump]
max_allowed_packet = 32M

[mysql]
no-auto-rehash

----------
with mysqld started with this config, @@slow_query_log is set to 1, @@global.long_query_time is set to 0.5, and slow_query_log_use_global_control is set t o'', but the server keeps sending all queries to the slow log.

here are the relevant variables when this happens:

mysql> select @@long_query_time\G
*************************** 1. row ***************************
@@long_query_time: 0.500000
1 row in set (0.00 sec)

mysql> show global variables like '%slow%'\G
*************************** 1. row ***************************
Variable_name: log_slow_admin_statements
Value: OFF
*************************** 2. row ***************************
Variable_name: log_slow_filter
Value:
*************************** 3. row ***************************
Variable_name: log_slow_queries
Value: OFF
*************************** 4. row ***************************
Variable_name: log_slow_rate_limit
Value: 1
*************************** 5. row ***************************
Variable_name: log_slow_rate_type
Value: query
*************************** 6. row ***************************
Variable_name: log_slow_slave_statements
Value: OFF
*************************** 7. row ***************************
Variable_name: log_slow_sp_statements
Value: OFF
*************************** 8. row ***************************
Variable_name: log_slow_verbosity
Value: microtime,query_plan,innodb
*************************** 9. row ***************************
Variable_name: slow_launch_time
Value: 2
*************************** 10. row ***************************
Variable_name: slow_query_log
Value: OFF
*************************** 11. row ***************************
Variable_name: slow_query_log_file
Value: /var/log/mysql/slowquery.log
*************************** 12. row ***************************
Variable_name: slow_query_log_timestamp_always
Value: ON
*************************** 13. row ***************************
Variable_name: slow_query_log_timestamp_precision
Value: microsecond
*************************** 14. row ***************************
Variable_name: slow_query_log_use_global_control
Value:
14 rows in set (0.00 sec)

Environment

None

Smart Checklist

Activity

lpjirasync January 24, 2018 at 5:52 AM

**Comment from Launchpad by: Launchpad Janitor on: 03-08-2013 04:17:31

[Expired for Percona Server because there has been no activity for 60 days.]

lpjirasync January 24, 2018 at 5:52 AM

**Comment from Launchpad by: Laurynas Biveinis on: 03-06-2013 09:58:21

5.5.27 is quite old release and there have been slow log-related fixes in 5.5.29-30.0 even though there are still some bugs remaining to this day. Thus please check with that or ideally with the current release. I cannot guarantee but I guess that this bug is not platform-specific.

Done

Details

Assignee

Reporter

Priority

Smart Checklist

Created January 24, 2018 at 5:52 AM
Updated January 24, 2018 at 5:52 AM
Resolved January 24, 2018 at 5:52 AM