mysqld got signal 11

问题发生背景

问题实例之前使用的是percona server，是安装pmm镜像自带的数据库，之后通过mysqldump迁移到了MySQL server，目前是只有有pmm server 访问pmm库，MySQL server就crash，并通过守护进程mysqld_safe 自动重启，crash 集尝试重启的一部分信息如下

:: UTC - mysqld got signal  ;
This could be because you hit a bug. It is also possible that this binary
or one of the libraries it was linked against is corrupt, improperly built,
or misconfigured. This error can also be caused by malfunctioning hardware.
Attempting to collect some information that could help diagnose the problem.
As this is a crash and something is definitely wrong, the information
collection process might fail.

key_buffer_size=
read_buffer_size=
max_used_connections=
max_threads=
thread_count=
connection_count=
It is possible that mysqld could use up to
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads =  K  bytes of memory
Hope that's ok; if not, decrease some variables in the equation.

Thread pointer: 0x7f7ac8000b30
Attempting backtrace. You can use the following information to find out
where mysqld died. If you see no messages after this, something went
terribly wrong...
stack_bottom = 7f7b20769e28 thread_stack 0x40000
/usr/local/mysql/bin/mysqld(my_print_stacktrace+0x35)[0xf32bd5]
/usr/local/mysql/bin/mysqld(handle_fatal_signal+0x4a4)[0x77b844]
/lib64/libpthread.so.[0x3db380f710]
/usr/local/mysql/bin/mysqld(_ZN10Field_blob15copy_blob_valueEP11st_mem_root+0x30)[0x7ab940]
/usr/local/mysql/bin/mysqld(_Z25mysql_prepare_blob_valuesP3THDR4ListI4ItemEP11st_mem_root+0x35e)[0xe8309e]
/usr/local/mysql/bin/mysqld(_Z12write_recordP3THDP5TABLEP9COPY_INFOS4_+0x251)[0xe835d1]
/usr/local/mysql/bin/mysqld(_ZN14Sql_cmd_insert12mysql_insertEP3THDP10TABLE_LIST+0x8d2)[0xe84882]
/usr/local/mysql/bin/mysqld(_ZN14Sql_cmd_insert7executeEP3THD+0xce)[0xe8500e]
/usr/local/mysql/bin/mysqld(_Z21mysql_execute_commandP3THDb+0xd92)[0xd10fe2]
/usr/local/mysql/bin/mysqld(_ZN18Prepared_statement7executeEP6Stringb+0x320)[0xd3ddd0]
/usr/local/mysql/bin/mysqld(_ZN18Prepared_statement12execute_loopEP6StringbPhS2_+0xdb)[0xd3e04b]
/usr/local/mysql/bin/mysqld(_Z19mysqld_stmt_executeP3THDmmPhm+0x11f)[0xd3e4cf]
/usr/local/mysql/bin/mysqld(_Z16dispatch_commandP3THDPK8COM_DATA19enum_server_command+0x74c)[0xd15c3c]
/usr/local/mysql/bin/mysqld(_Z10do_commandP3THD+0x194)[0xd17114]
/usr/local/mysql/bin/mysqld(handle_connection+0x294)[0xde1c34]
/usr/local/mysql/bin/mysqld(pfs_spawn_thread+0x174)[0x11f96f4]
/lib64/libpthread.so.[0x3db38079d1]
/lib64/libpthread.so.[0x3db38079d1]
/lib64/libc.so.(clone+0x6d)[0x3db34e8b6d]

Trying to get some variables.
Some pointers may be invalid and cause the dump to abort.
Query (7f7ac80054f0): Query (7f7ac80054f0): is an invalid pointer
Connection ID (thread ID):
Status: NOT_KILLED

The manual page at http://dev.mysql.com/doc/mysql/en/crashing.html contains
information that should help you find out what is causing the crash. :: mysqld_safe Number of processes running now:  :: mysqld_safe mysqld restarted
--11T11::.827511Z  [Warning] Insecure configuration for --secure-file-priv: Current value does not restrict location of generated files. Consider setting it to a valid, non-empty path.

原 percona server 版本：5.7.18-16；目前MySQL server 版本：5.7.10

尝试将MySQL server 版本升级到5.7.20，无法解决问题，后来变更为percona server 5.7的一个分支版本解决问题

通过rmp 安装percona server

官方文档

# 安装yum 源
yum install http://www.percona.com/downloads/percona-release/redhat/0.1-4/percona-release-0.1-4.noarch.rpm
# 安装percona server
yum install Percona-Server-server-

防止使用my.cnf启动，指定配置文件启动服务，使用现有的配置文件和数据文件

# 启动服务
/usr/bin/mysqld_multi --defaults-file=/etc/mysql/.cnf --mysqladmin=/usr/bin/mysqladmin --mysqld=/usr/bin/mysqld_safe --no-log start 3306
# 关闭服务
/usr/bin/mysqld_multi --defaults-file=/etc/mysql/3306.cnf --mysqladmin=/usr/bin/mysqladmin --mysqld=/usr/bin/mysqld_safe --no-log stop 3306
# 登陆
/usr/bin/mysql -S /tmp/mysql_3306.sock

关于MySQL server的pid-file 和mysqld_multi 的启动项

由于启动的选项是--defaults-extra-file 所以同时会读取全局变量文件/etc/my.cnf，可以设置为

[mysqld]
...
[mysqld_safe]
...
[mysqld_multi]

mysqld=/usr/local/mysql/bin/mysqld_safe
mysqladmin=/usr/local/mysql/bin/mysqladmin

对于--mysqld 的值可以是二进制文件mysqld的位置（会读取[mysqld]下的配置），守护进程mysqld_safe不会启动

也可以是mysqld_safe启动脚本的位置（会读取[mysqld_safe]下的配置），mysqld和守护进程mysqld_safe都会启动，如下可以看到两者的运行信息

ps -ef|grep
root             : pts/    :: /bin/sh /usr/local/mysql/bin/mysqld_safe --server-id= --user=mysql --port= --datadir=/data0/mysql/3309_test --socket=/tmp/mysql_3309.sock --innodb_data_home_dir=/data0/mysql/3309_test --innodb_log_group_home_dir=/data0/mysql/3309_test --transaction-isolation=READ-COMMITTED --explicit_defaults_for_timestamp=on --max_allowed_packet=100M --read_rnd_buffer_size=8M --tmp_table_size=128M --thread_cache_size= --max_connections= --max_heap_table_size=128M --open-files-limit= --max_connect_errors= --symbolic-links= --tmpdir=/data0/mysql/tmp --innodb_online_alter_log_max_size=1G --innodb_data_file_path=ibdata1:512M:autoextend --innodb_log_files_in_group= --innodb_buffer_pool_size=256M --innodb_log_file_size=100M --innodb_flush_log_at_trx_commit= --innodb_open_files= --innodb_lock_wait_timeout= --log-error=log-error --slow_query_log_file=slow_query.log --slow_launch_time= --slow_query_log= --log-bin=mysql-bin --binlog_format=ROW --max_binlog_size=1G --expire_logs_days= --master-info-repository=TABLE --relay-log-info-repository=TABLE --gtid-mode=on --enforce-gtid-consistency=on --log-slave-updates= --relay-log=relay-log --sync-master-info= --skip-slave-start --slave_pending_jobs_size_max= --log_bin_trust_function_creators= --report-port= --report-host=10.211.253.192 --slave-sql-verify-checksum= --master-verify-checksum= --slave_parallel_type=LOGICAL_CLOCK --slave-parallel-workers= --binlog_group_commit_sync_delay= --skip-name-resolve --lower_case_table_names= --character-set-server=utf8 --sql_mode=NO_AUTO_VALUE_ON_ZERO,STRICT_TRANS_TABLES,NO_AUTO_CREATE_USER,NO_ENGINE_SUBSTITUTION --slave_preserve_commit_order= --default_password_lifetime= --binlog_rows_query_log_events=
mysql        : pts/    :: /usr/local/mysql/bin/mysqld --basedir=/usr/local/mysql --datadir=/data0/mysql/3309_test --plugin-dir=/usr/local/mysql/lib/plugin --user=mysql --server-id= --innodb-data-home-dir=/data0/mysql/3309_test --innodb-log-group-home-dir=/data0/mysql/3309_test --transaction-isolation=READ-COMMITTED --explicit-defaults-for-timestamp=on --max-allowed-packet=100M --read-rnd-buffer-size=8M --tmp-table-size=128M --thread-cache-size= --max-connections= --max-heap-table-size=128M --max-connect-errors= --symbolic-links= --tmpdir=/data0/mysql/tmp --innodb-online-alter-log-max-size=1G --innodb-data-file-path=ibdata1:512M:autoextend --innodb-log-files-in-group= --innodb-buffer-pool-size=256M --innodb-log-file-size=100M --innodb-flush-log-at-trx-commit= --innodb-open-files= --innodb-lock-wait-timeout= --slow-query-log-file=slow_query.log --slow-launch-time= --slow-query-log= --log-bin=mysql-bin --binlog-format=ROW --max-binlog-size=1G --expire-logs-days= --master-info-repository=TABLE --relay-log-info-repository=TABLE --gtid-mode=on --enforce-gtid-consistency=on --log-slave-updates= --relay-log=relay-log --sync-master-info= --skip-slave-start --slave-pending-jobs-size-max= --log-bin-trust-function-creators= --report-port= --report-host=10.211.253.192 --slave-sql-verify-checksum= --master-verify-checksum= --slave-parallel-type=LOGICAL_CLOCK --slave-parallel-workers= --binlog-group-commit-sync-delay= --skip-name-resolve --lower-case-table-names= --character-set-server=utf8 --sql-mode=NO_AUTO_VALUE_ON_ZERO,STRICT_TRANS_TABLES,NO_AUTO_CREATE_USER,NO_ENGINE_SUBSTITUTION --slave-preserve-commit-order= --default-password-lifetime= --binlog-rows-query-log-events= --log-error=/data0/mysql/3309_test/log-error.err --open-files-limit= --pid-file=/data0/mysql/3309_test/mysql-bcc03.dev.yz.pid --socket=/tmp/mysql_3309.sock --port=

如果pid-file在配置文件中没有指定，则默认为$datadir/`hostname`.pid