Details
Description
Hi,
my server has wait_timeout set to 28800 seconds (8 hours).
Looking at my sessions idling ...
MariaDB [(none)]> select * FROM INFORMATION_SCHEMA.PROCESSLIST ORDER BY TIME DESC LIMIT 10; +-------+-------------+----------------------+--------------+---------+-------+--------------------+------+--------------+-------+-----------+----------+-------------+---------------+----------+ | ID | USER | HOST | DB | COMMAND | TIME | STATE | INFO | TIME_MS | STAGE | MAX_STAGE | PROGRESS | MEMORY_USED | EXAMINED_ROWS | QUERY_ID | +-------+-------------+----------------------+--------------+---------+-------+--------------------+------+--------------+-------+-----------+----------+-------------+---------------+----------+ | 2 | system user | | NULL | Sleep | 43731 | NULL | NULL | 43731770.086 | 0 | 0 | 0.000 | 35296 | 0 | 0 | | 1 | system user | | NULL | Sleep | 43731 | wsrep aborter idle | NULL | 43731770.096 | 0 | 0 | 0.000 | 35296 | 0 | 0 | (...) | 73945 | webapp-xx | 205.xxx.xx.xxx:12345 | database_name | Sleep | 28590 | | NULL | 28590778.291 | 0 | 0 | 0.000 | 50640 | 0 | 8197323 | (...)
Then in the log, I can see:
150313 11:18:20 [Warning] Aborted connection 73645 to db: 'database_name user: 'webapp-xx' host: '205.xxx.xx.xxx' (Unknown error)
I think the error message should be more explicit, such as "connection closed due to wait_timeout limit reached".
To reproduce, start mariadb server, with --log_warnings=2, set wait_timeout, and let application session idle until the timeout, then look at error log.
Thanks,
Joffrey
Gliffy Diagrams
Attachments
Activity
- All
- Comments
- Work Log
- History
- Activity
- Transitions
It is actually a regression. We used to have a proper message in 5.1-5.3, but lost it in 5.5 and higher. MySQL 5.5 produces the correct message.