* Password verification and v-for add key
* DAG automatic layout
* combo.js replacement
* combo.js replacement
* webpack upgrade to 4
* packjson has a file name that has not changed
* fix issue:1477 some tasks would be running all the time when db delayed
* fix issue:1477 some tasks would be running all the time when db delayed
* fix issue:1477 some tasks would be running all the time when db delayed
* add Spark Version in Spark Component
add Spark Version in Spark Component
* add license for SparkVersion.class
add license
* 1 add spark task UT
2 add spark version param check
* add assert check for sparkTaskTest
* fix:When there are multiple nodes, the monitoring center dashboard displays only one node.
* fix:Catching exceptions and printing exception information
* fix:rollback
* fix:getting project name field bug
To improve user experience in datasource page, add default port for each type database.
1. When create or edit datasource, there is a default vaule for port. If user changes type, the default value changes.
2. Use local variables in browser memory, to cache user previous input port for each type. For example :
time-a: MySQL default port is 3306 , user change 3306 to 4000.
time-b: User change type from MySQL to ClickHouse , the port change from 4000 to 8123.
time-c: User change type from ClickHouse back to MySQL,the port change from 8123 to 4000,use previous input cache to fillback.
3. In edit datasource page, when the page first loads, use the query value fillback instead of default value.
Now all datatype use same method ` stmt.setString(index,value);` even DATE and TIMESTAMP.
So i had improved it. If datatype is DATE use `stmt.setDate(index, java.sql.Date.valueOf(value));` , if datatype is TIMESTAMP use `stmt.setTimestamp(index, java.sql.Timestamp.valueOf(value));`.i
When use DATE the param format must use $[yyyy-MM-dd] , When use TIMESTAMP the param format must use $[yyyy-MM-dd hh:mm:ss].
public static final String JDBC_POSTGRESQL_CLASS_NAME = "org.postgresql.Driver";
and
public static final String ORG_POSTGRESQL_DRIVER = "org.postgresql.Driver";
There are same effect, We can only use one of them.
* [dolphinscheduler-1345] [newfeature] Add DB2 Datasource (#1391)
* Fix the problem that the 'queueId' is not present when creating a tenant based on the default queue. (#1409)
* dolphinscheduler-#1403][bug]improve the check rules (#1408)
1. When check failed, we don’t know whitch parameter has is wrong, Because username、password、email and phone were checks together. I refactored the check method ,Now it will return failed msg by each field.
2. The email check regex support [_|\-|\.]?) in createUser.vue, But it do not support in backend server , I fix it, Now they have the same check regex both in frontend and backend
* jcip-annotations define version information and maven-assembly-plugin add groupId (#1413)
* "v-for" add key (#1419)
* [dolphinscheduler-#1397] [bug]Resources can not be previewed or updated (#1406)
When create an resource the name will add the suffix, But When rename the resource there is no suffix add to the name, So When update resource name without suffix just like "test.sh" => "test" , Then the bug reproduced.
To fix this bug i add the logic bellow:
When rename, if the name without suffix then add it ,else use the origin name
* simply server module configs (#1424)
* move updateTaskState into try/catch block in case of exception
* fix NPE
* using conf.getInt instead of getString
* for AbstractZKClient, remove the log, for it will print the same log message in createZNodePath.
for AlertDao, correct the spelling.
* duplicate
* refactor getTaskWorkerGroupId
* add friendly log
* update hearbeat thread num = 1
* fix the bug when worker execute task using queue. and remove checking Tenant user anymore in TaskScheduleThread
* 1. move verifyTaskInstanceIsNull after taskInstance
2. keep verifyTenantIsNull/verifyTaskInstanceIsNull clean and readable
* fix the message
* delete before check to avoid KeeperException$NoNodeException
* fix the message
* check processInstance state before delete tenant
* check processInstance state before delete worker group
* refactor
* merge api constants into common constatns
* update the resource perm
* update the dataSource perm
* fix CheckUtils.checkUserParams method
* update AlertGroupService, extends from BaseService, remove duplicate methods
* refactor
* modify method name
* add hasProjectAndPerm method
* using checkProject instead of getResultStatus
* delete checkAuth method, using hasProjectAndPerm instead.
* correct spelling
* add transactional for deleteWorkerGroupById
* add Transactional for deleteProcessInstanceById method
* change sqlSessionTemplate singleton
* change sqlSessionTemplate singleton and reformat code
* fix unsuitable error message
* update shutdownhook methods
* fix worker log bug
* fix api server debug mode bug
* upgrade zk version
* delete this line ,for zkClient.close() will do the whole thing
* fix master server shutdown error
* degrade zk version and add FourLetterWordMain class
* fix PathChildrenCache not close
* add Transactional for createSession method
* add more message for java-doc
* delete App, let spring manage connectionFactory
* add license
* add class Application for test support
* refactor masterServer and workerServer
* add args
* fix the spring transaction not work bug
* remove author
* delete @Bean annotation
* delete master/worker properties
* updates
* rename application.properties to application-dao.properties
* delete this class
* delete master/worker properties and refactory master/worker
* delete unused imports
* merge
* delete unused config
When workgroup referenced by workflow , And there isn’t any instace in
workflow ,the workgroup can be delete . After the workgroup delete , Use the
default workgroup in the page ,Instead of the value id.
* Solve the problem that the pop-up box cannot be closed when the browser returns
* Change the ans-ui version and API interface name
* Add page number to list page