* It's highly recommended to share one stateless handler with all the server-side channels.
* comment for prepareSharableHandlers
* Simplify comment
* remove comment
* supply test case
* add logic of putting message that exceeds max-check-times to system topic TRANS_CHECK_MAXTIME_TOPIC
* add test case:testResolveDiscardMsg
* add @After logic to test case
* comment brokerController.shutdown and use mock
* add logic of resuming half message check
* add test case:resumeCheckHalfMessage
* delete commented codes
* [issue#1164]return the codes to original reput method part.
* [issue#1164]fix issue that Consumer Instance can't consume message from slave when cluster is in the high level tps and master has been killed.
* [issue#1164]if the broker is a master node,then modify reputFromOffset correctly.
* [issue#1164]add some coding comments.
* [ISSUE#1156]new mqadmin API for ACL configuration.Add dataVersion and write data to acl yaml config file in the acl module.
* [ISSUE#1156]add unit test cases for new acl mqadmin API command which adding dataversion part.
* [ISSUE#1156]implement update,delete and query Acl config version in new acl mqadmin API command.
* [ISSUE#1156]polish and optimize the implementation for new acl mqadmin API command.
* [ISSUE#1156]fix the issues that topicPerms and groupPerms can't be updated and there is NPE when querying acl config version.
* [ISSUE#1156]fix small issue that specify wrong plainAccessConfig attribute,the correct one is AccessKey.
* [ISSUE#1156]add updateGlobalWhiteAddr subcommand codes for mqadmin acl command.
* [ISSUE#1156]adjust some codes for cluster acl config version list in the mqadmin acl commands.
* [ISSUE#1156]add acl mqadmin command part in the acl user_guide docs.
* [ISSUE#1156]polish and optimize some part of acl mqadmin command codes.
* [ISSUE#1156]fix code comments issue that the first letter needs to be capitalized and adjust the contents of acl user_guide.
huangli