MariaDB 10.11.4出现innodb_fatal_semaphore_wait_threshold超阈值问题求助

Debian Bookworm系统上的MariaDB Server 10.11.4频繁触发致命错误：InnoDB: innodb_fatal_semaphore_wait_threshold was exceeded for dict_sys.latch。此前未修改底层系统及数据库配置，已将innodb_buffer_pool_size从4GB调整至40GB，但问题仍未解决。相关错误日志如下：

mariadbd[324032]: 2023-12-22 11:50:54 0 [ERROR] [FATAL] InnoDB: 
innodb_fatal_semaphore_wait_threshold was exceeded for dict_sys.latch. 
Please refer to
https://mariadb.com/kb/en/how-to-produce-a-full-stack-trace-for-mysqld/

mariadbd[324032]: 231222 11:50:54 [ERROR] mysqld got signal 6 ;
mariadbd[324032]: This could be because you hit a bug. It is also possible that this binary
mariadbd[324032]: or one of the libraries it was linked against is corrupt, improperly built,
mariadbd[324032]: or misconfigured. This error can also be caused by malfunctioning hardware.
mariadbd[324032]: To report this bug, see https://mariadb.com/kb/en/reporting-bugs
mariadbd[324032]: We will try our best to scrape up some info that will hopefully help
mariadbd[324032]: diagnose the problem, but since we have already crashed,
mariadbd[324032]: something is definitely wrong and this may fail.
mariadbd[324032]: Server version: 10.11.4-MariaDB-1~deb12u1 source revision:
mariadbd[324032]: key_buffer_size=134217728
mariadbd[324032]: read_buffer_size=131072
mariadbd[324032]: max_used_connections=99
mariadbd[324032]: max_threads=1002
mariadbd[324032]: thread_count=99
mariadbd[324032]: It is possible that mysqld could use up to
mariadbd[324032]: key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 2337911 K  bytes of memory
mariadbd[324032]: Hope that's ok; if not, decrease some variables in the equation.
mariadbd[324032]: Thread pointer: 0x0
mariadbd[324032]: Attempting backtrace. You can use the following information to find out
mariadbd[324032]: where mysqld died. If you see no messages after this, something went
mariadbd[324032]: terribly wrong...
mariadbd[324032]: stack_bottom = 0x0 thread_stack 0x49000
mariadbd[324032]: 
/usr/sbin/mariadbd(my_print_stacktrace+0x2e)[0x55ea6dcb14ae]
mariadbd[324032]: 
/usr/sbin/mariadbd(handle_fatal_signal+0x409)[0x55ea6d81ef49]
mariadbd[324032]: /lib/x86_64-linux-gnu/libc.so.6(+0x3bfd0)[0x7f9bb5e5afd0]
mariadbd[324032]: /lib/x86_64-linux-gnu/libc.so.6(+0x8ad3c)[0x7f9bb5ea9d3c]
mariadbd[324032]: 
/lib/x86_64-linux-gnu/libc.so.6(gsignal+0x12)[0x7f9bb5e5af32]
mariadbd[324032]: 
/lib/x86_64-linux-gnu/libc.so.6(abort+0xd3)[0x7f9bb5e45472]
mariadbd[324032]: /usr/sbin/mariadbd(+0x68b8af)[0x55ea6d4818af]
mariadbd[324032]: /usr/sbin/mariadbd(+0x683b58)[0x55ea6d479b58]
mariadbd[324032]: 
/usr/sbin/mariadbd(_ZN5tpool19thread_pool_generic13timer_generic7executeEPv+0x38)[0x55ea6dc555a8]
mariadbd[324032]: 
/usr/sbin/mariadbd(_ZN5tpool4task7executeEv+0x2f)[0x55ea6dc563ef]
mariadbd[324032]: 
/usr/sbin/mariadbd(_ZN5tpool19thread_pool_generic11worker_mainEPNS_11worker_dataE+0x4f)[0x55ea6dc5466f]
mariadbd[324032]: 
/lib/x86_64-linux-gnu/libstdc++.so.6(+0xd44a3)[0x7f9bb60d44a3]
mariadbd[324032]: /lib/x86_64-linux-gnu/libc.so.6(+0x89044)[0x7f9bb5ea8044]
mariadbd[324032]: /lib/x86_64-linux-gnu/libc.so.6(+0x10961c)[0x7f9bb5f2861c]
mariadbd[324032]: The manual page at
https://mariadb.com/kb/en/how-to-produce-a-full-stack-trace-for-mysqld/
contains
mariadbd[324032]: information that should help you find out what is causing the crash.
mariadbd[324032]: Writing a core file...
mariadbd[324032]: Working directory at /var/lib/mysql
mariadbd[324032]: Resource Limits:
mariadbd[324032]: Limit                     Soft Limit Hard Limit           Units
mariadbd[324032]: Max cpu time              unlimited unlimited            seconds
mariadbd[324032]: Max file size             unlimited unlimited            bytes
mariadbd[324032]: Max data size             unlimited unlimited            bytes
mariadbd[324032]: Max stack size            8388608 unlimited bytes
mariadbd[324032]: Max core file size        0 unlimited            bytes
mariadbd[324032]: Max resident set          unlimited unlimited            bytes
mariadbd[324032]: Max processes             1030903 1030903 processes
mariadbd[324032]: Max open files            200000 200000 files
mariadbd[324032]: Max locked memory         524288 524288 bytes
mariadbd[324032]: Max address space         unlimited unlimited            bytes
mariadbd[324032]: Max file locks            unlimited unlimited            locks
mariadbd[324032]: Max pending signals       1030903 1030903 signals
mariadbd[324032]: Max msgqueue size         819200 819200 bytes
mariadbd[324032]: Max nice priority         0                    0
mariadbd[324032]: Max realtime priority     0                    0
mariadbd[324032]: Max realtime timeout      unlimited unlimited            us
mariadbd[324032]: Core pattern: core
mariadbd[324032]: Kernel version: Linux version 6.1.0-12-amd64
(debian-kernel@lists.debian.org) (gcc-12 (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40) #1 SMP PREEMPT_DYNAMIC Debian
6.1.52-1 (2023-09-07)
systemd[1]: mariadb.service: Main process exited, code=killed, status=6/ABRT
systemd[1]: mariadb.service: Failed with result 'signal'.
systemd[1]: mariadb.service: Consumed 3h 50min 55.933s CPU time.
systemd[1]: mariadb.service: Scheduled restart job, restart counter is at 1.
systemd[1]: Stopped mariadb.service - MariaDB 10.11.4 database server.
systemd[1]: mariadb.service: Consumed 3h 50min 55.933s CPU time.
systemd[1]: Starting mariadb.service - MariaDB 10.11.4 database server...

• InnoDB字典锁长时间竞争：dict_sys.latch是InnoDB保护数据字典（表结构、索引元数据）的核心锁。当大量线程同时请求该锁，且某个线程持有锁的时间超过innodb_fatal_semaphore_wait_threshold阈值（默认600秒）时，就会触发这个致命错误。常见场景包括频繁执行DDL操作，或查询密集访问元数据导致锁竞争加剧。
• MariaDB 10.11.4版本bug：从堆栈信息看，崩溃发生在线程池定时器执行逻辑中，该版本可能存在线程池调度与InnoDB锁机制交互的bug，导致锁等待超时触发崩溃。
• 系统硬件或资源瓶颈：即使资源限制显示正常，也可能存在磁盘I/O响应缓慢（导致锁持有时间延长）、内存交换频繁，或CPU被其他进程耗尽，使得InnoDB线程无法及时释放锁。
• 依赖库或二进制文件问题：MariaDB依赖的系统库（如libc、libstdc++）可能损坏，或者MariaDB二进制包编译存在缺陷，导致锁逻辑执行异常。

内容的提问来源于stack exchange，提问作者peter170567