hcq/MaxScale
hcq/MaxScale
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix mxs359_read_only

Browse Source 
Due to the changes in the monitor, an explicit failcount=1 and extra waits
are required to make sure the master actually changes.
This commit is contained in:
Markus Mäkelä 2018-11-09 09:26:42 +02:00
parent a9e2364979
commit 514b5f7856
No known key found for this signature in database
GPG Key ID: 72D48FCE664F7B19
2 changed files with 2 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
maxscale-system-test/cnf/maxscale.cnf.template.mxs359_read_only
View File

@ -8,6 +8,7 @@ servers=server1,server2,server3,server4
user=maxskysql
password=skysql
monitor_interval=1000
failcount=1
[RW-Split-Router]
type=service

2
maxscale-system-test/mxs359_read_only.cpp
View File

@ -65,7 +65,7 @@ void test_new_master(TestConnections& test, std::ostream& out)
test.try_query(test.maxscales->conn_rwsplit[0], "SELECT * FROM test.t1");
change_master(1, 0);
test.maxscales->wait_for_monitor();
test.maxscales->wait_for_monitor(2);
out << "Both reads and writes after master change should work" << endl;
test.try_query(test.maxscales->conn_rwsplit[0], "INSERT INTO test.t1 VALUES (2)");