Comments - error : (408595) Lost connection to the master server

5 years, 1 month ago Markus Mäkelä

The idle value for all of those was around 300 seconds, what is the wait_timeout value configured on the backend servers? That error is likely caused by the server closing idle connections and the master connection was not used in the last 300 seconds.

You can set the value of connection_keepalive for the readwritesplit service to 60 seconds to keep the connection alive:

connection_keepalive=60
 
5 years, 1 month ago mi saki

it's not work

+-----------------------------+----------+
| Variable_name               | Value    |
+-----------------------------+----------+
| connect_timeout             | 30       |
| deadlock_timeout_long       | 50000000 |
| deadlock_timeout_short      | 10000    |
| delayed_insert_timeout      | 300      |
| innodb_flush_log_at_timeout | 1        |
| innodb_lock_wait_timeout    | 120      |
| innodb_rollback_on_timeout  | OFF      |
| interactive_timeout         | 300      |
| lock_wait_timeout           | 86400    |
| net_read_timeout            | 30       |
| net_write_timeout           | 60       |
| slave_net_timeout           | 60       |
| thread_pool_idle_timeout    | 60       |
| wait_timeout                | 300      |
+-----------------------------+----------+
14 rows in set (0.00 sec)

 
5 years, 1 month ago Markus Mäkelä

Try increasing it to a higher value:

SET GLOBAL wait_timeout=3600;

That should allow MaxScale enough time to keep the connection alive with the default connection_keepalive of 300 seconds.

 
Content reproduced on this site is the property of its respective owners, and this content is not reviewed in advance by MariaDB. The views, information and opinions expressed by this content do not necessarily represent those of MariaDB or any other party.