[MDEV-24091] Galera crashes with "WSREP: FSM: no such a transition REPLICATING -> COMMITTED" Created: 2020-11-02  Updated: 2023-10-31  Resolved: 2021-11-18

Status: Closed
Project: MariaDB Server
Component/s: Galera
Affects Version/s: 10.4.14
Fix Version/s: N/A

Type: Bug Priority: Critical
Reporter: Valerii Kravchuk Assignee: Seppo Jaakola
Resolution: Cannot Reproduce Votes: 1
Labels: need_feedback

Issue Links:
Relates
relates to MDEV-32561 WSREP FSM failure: no such a transiti... Closed

 Description   

MariaDB crashes with the following messages:

2020-11-02 15:51:10 682745 [ERROR] WSREP: FSM: no such a transition REPLICATING -> COMMITTED
 
201102 15:51:10 [ERROR] mysqld got signal 6 ;
 
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.
 
 
 
To report this bug, see https://mariadb.com/kb/en/reporting-bugs
 
 
 
We will try our best to scrape up some info that will hopefully help
 
diagnose the problem, but since we have already crashed, 
something is definitely wrong and this may fail.
 
 
 
Server version: 10.4.14-MariaDB-log
 
key_buffer_size=33554432
 
read_buffer_size=2097152
 
max_used_connections=737
 
max_threads=4011
 
thread_count=700
 
It is possible that mysqld could use up to 
key_buffer_size + (read_buffer_size + sort_buffer_size)*max_threads = 41352608 K  bytes of memory
 
Hope that's ok; if not, decrease some variables in the equation.
 
 
 
Thread pointer: 0x7fcbfcfa1ec8
 
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 = 0x7fd944857e40 thread_stack 0x49000
 
*** stack smashing detected ***: /usr/local/mysql/bin/mysqld terminated
 
Fatal signal 11 while backtracing

Unfortunately crash during backtrace gives us no further details.

 Comments   
Comment by Jan Lindström (Inactive) [ 2021-01-12 ]

I would need a repeatable test case or full stack dump. Both stack dumps here are not full and contain either nothing or strange addresses only e.g. in last one 0x0 ??

Comment by Seppo Jaakola [ 2021-01-27 ]

Looking at the stack traces, it seems that the crash would happen with a scenario where:

  • DML (INSERT in one of the stack traces), is executed as prepared statement (application is using binary protocol prepared statement)
  • The target table has a trigger defined, and the trigger code has a stored procedure call. The trigger fires and the stored procedure is called
  • The stored procedure calls further another stored procedure (this may not be important for reproducing the crash)
  • The stored procedure execution is terminated by high priority incoming replication, which conflicts with the actual DML or operations carried out by the stored procedure execution
Generated at Thu Feb 08 09:27:22 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.