Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
2.5.9
-
None
-
Ubuntu 18.04.5 & RHEL 8.3
Description
MASTER_GTID_WAIT accept timeout value from 0.000001s and maxscale 2.5 has no problem starting with value lower than 1s compared to 2.4.
Mariadb general_log:
SET @maxscale_secret_variable=(SELECT CASE WHEN MASTER_GTID_WAIT('1-13-474756993', 0) = 0 THEN 1 ELSE (SELECT 1 FROM INFORMATION_SCHEMA.ENGINES) END);SELECT ..
From maxctrl show service Splitter-Service
...
"causal_reads_timeout": 500,
...
conf:
[maxscale]
|
threads=8
|
syslog=1
|
maxlog=0
|
log_notice=0
|
load_persisted_configs=false
|
local_address=10.1.1.23
|
auth_connect_timeout=20s
|
users_refresh_time=10s
|
max_auth_errors_until_block=100
|
admin_gui=0
|
|
|
[master]
|
type=server
|
address=10.1.1.121
|
port=3306
|
protocol=MariaDBBackend
|
|
|
[slave-1]
|
type=server
|
address=10.1.1.14
|
port=3306
|
protocol=MariaDBBackend
|
|
|
[Monitor]
|
type=monitor
|
module=mariadbmon
|
servers=master,slave-1
|
user=xxx
|
password=xxx
|
monitor_interval=3s
|
assume_unique_hostnames=OFF
|
master_conditions=none
|
slave_conditions=linked_master
|
ignore_external_masters=ON
|
|
|
[Splitter-Service]
|
type=service
|
router=readwritesplit
|
servers=master,slave-1
|
user=xxx
|
password=xxx
|
#session_track_trx_state=1
|
max_slave_replication_lag=5s
|
max_sescmd_history=500
|
slave_connections=1
|
master_accept_reads=true
|
use_sql_variables_in=master
|
master_failure_mode=fail_on_write
|
causal_reads=local
|
causal_reads_timeout=500ms
|
|
|
[ReadWrite-Listener]
|
type=listener
|
service=Splitter-Service
|
protocol=MariaDBClient
|
socket=/run/maxscale/maxscale.sock
|