mo-tester

This directory contains a test suite for the MatrixOne engine.

How to use?

This tester is designed to test MatrixOne or any other database functionalities with SQL.

0. Install JDK8 if you don't have it yet.

1: Run your MatrixOne instance or other DB instance.

Checkout Install MatrixOne to launch a MatrixOne instance.

Or you can launch whatever database software as you want.

2. Fork and clone this mo-tester project.

3. Configure the connection.

In mo.yml file, configure the server address, default database name, username&password etc. MO-tester is based on java, so these parameters are required for JDBC driver. Below is a default example for a local standalone version MatrixOne.

#jdbc
jdbc:
  driver: "com.mysql.cj.jdbc.Driver"
  server:
  - addr: "127.0.0.1:6001"
  database:
    default: "test"
  paremeter:
    characterSetResults: "utf8"
    continueBatchOnError: "false"
    useServerPrepStmts: "true"
    alwaysSendSetIsolation: "false"
    useLocalSessionState: "true"
    zeroDateTimeBehavior: "CONVERT_TO_NULL"
    failoverReadOnly: "false"
    serverTimezone: "Asia/Shanghai"

#users
user:
  name: "dump"
  passwrod: "111"

4. Run the test.

With the simple below command, all the SQL test cases will automatically run and generate reports and error messages to report/report.txt and report/error.txt.

> ./run.sh

If you'd like to adjust the test range, you can just change the path parameter of run.yml.

And you can also specify some parameters when executing the command run.sh,parameters are as followings:

-p  set the path of test cases needed to be executed by mo-tester,default value is configured by the `path` in `run.yaml`
-m  set the method that mo-tester will run with,default value is configured by the `method` in `run.yaml`
-t  set the type of the format that mo-tester execute the sqlcommand in,default value is configured by the `type` in `run.yaml`
-r  set The success rate that test cases should reach,default value is configured by the `rate` in `run.yaml`
-i  set the including list, and only script files in the path which name contain one of the list will be excuted,if more than one,seperated by `,`,if not specified,refers to all cases included
-e  set the excluding list, and script files in the path which name contain one of the list will not be excuted,if more than one,seperated by `,`,if not specified,refers to none of cases excluded
-g  means SQL commands which is marked with [bvt:issue] flag will not be executed,this flag starts with [-- @bvt:issue#{issueNO.}],and ends with [-- @bvt:issue],eg:
    -- @bvt:issue#3236
    select date_add("1997-12-31 23:59:59",INTERVAL "-10000:1" HOUR_MINUTE);
    select date_add("1997-12-31 23:59:59",INTERVAL "-100 1" YEAR_MONTH);
    -- @bvt:issue
    Those two sql commands are associated with the issue#3236,and they will not been executed in bvt test,until the flag is removed when the issue#3236 is fixed.
    
-n  means the meta data of the resultset will be ignored when comparing the result
Examples:
bash run.sh -p case -m run -t script -r 100 -i select,subquery -e substring -g

If you want to automatically generate SQL results for the new SQL cases, you can just change the method parameter of run.yml to genrs. Running the run.sh scripts will directly record test results in the result/ path with their original filenames.

Note: everytime running run.sh will overwrite the error, report and success reports.