|
I have configured readwrite in maxscale, and set the master_failure_mode to fail_on_write (error_on_write), but when I close the master of mysql, all connecion had been closed and any new session connect to maxscale listener port is failure, and logs in /var/log/maxscale/maxscale.log
stop master:
name: [234.008_5506_test] status: [Slave, Running] state: [NOT_IN_USE] last opened at: [not opened] last closed at: [not closed] last close reason: [] num sescmd: [0]
|
name: [234.008_5507_test] status: [Down] state: [NOT_IN_USE] last opened at: [not opened] last closed at: [not closed] last close reason: [] num sescmd: [0]
|
name: [234.008_5508_test] status: [Slave, Running] state: [NOT_IN_USE] last opened at: [not opened] last closed at: [not closed] last close reason: [] num sescmd: [0]
|
2021-12-05 22:28:00 error : (99491) [mariadbclient] Routing the query failed. Session will be closed.
|
2021-12-05 22:28:01 error : (99492) [readwritesplit] (orche_service) Could not find valid server for target type TARGET_SLAVE (COM_QUERY: show databases), closing connection.
|
orche_service.cnf
[orche_service]
|
type=service
|
router=readwritesplit
|
router_options=master,slave
|
user=maxscale
|
password=xxx
|
enable_root_user=false
|
max_connections=0
|
connection_timeout=0
|
net_write_timeout=0
|
auth_all_servers=false
|
strip_db_esc=true
|
log_auth_warnings=true
|
session_track_trx_state=false
|
retain_last_statements=-1
|
session_trace=false
|
rank=primary
|
connection_keepalive=300s
|
causal_reads=false
|
causal_reads_timeout=10000ms
|
delayed_retry=false
|
delayed_retry_timeout=10000ms
|
disable_sescmd_history=false
|
lazy_connect=false
|
master_accept_reads=false
|
master_failure_mode=fail_on_write
|
master_reconnection=true
|
max_sescmd_history=50
|
max_slave_connections=255
|
max_slave_replication_lag=6s
|
optimistic_trx=false
|
prune_sescmd_history=false
|
retry_failed_reads=true
|
slave_connections=255
|
slave_selection_criteria=LEAST_CURRENT_OPERATIONS
|
strict_multi_stmt=false
|
strict_sp_calls=false
|
transaction_replay=false
|
transaction_replay_attempts=5
|
transaction_replay_max_size=1073741824
|
transaction_replay_retry_on_deadlock=false
|
use_sql_variables_in=master
|
targets=234.008_5506_test,234.008_5507_test,234.008_5508_test
|
234.008_5508_test.cnf
[234.008_5508_test]
|
address=192.168.234.8
|
extra_port=0
|
persistmaxtime=0ms
|
persistpoolmax=0
|
port=5508
|
priority=0
|
proxy_protocol=false
|
rank=primary
|
ssl=false
|
ssl_cert_verify_depth=9
|
ssl_verify_peer_certificate=false
|
ssl_verify_peer_host=false
|
ssl_version=MAX
|
type=server
|
I used detect in 2.4 but they seems to be deprecated*
2021-12-05 22:26:35 warning: [mariadbmon] 'detect_stale_slave' is deprecated and should not be used. Use 'slave_conditions' instead.
2021-12-05 22:26:35 warning: [mariadbmon] 'detect_standalone_master' is deprecated and should not be used. Use 'master_conditions' instead.
2021-12-05 22:26:35 warning: [mariadbmon] 'detect_stale_master' is deprecated and should not be used. Use 'master_conditions' instead.
2021-12-05 22:26:35 warning: [mariadbmon] 'detect_stale_slave' is deprecated and should not be used. Use 'slave_conditions' instead.
Is this feature(when master is failure, the slaves are continue working) just take effect in Mariadb?
Are there any other paramters can be seted to support mysql?
|