XXL-JOB are competing for the "Year 2019 China's most popular open source software," look forward to your valuable vote! Polls Links
v2.1.1 Release Notes
- 1, the dispatch center logs automatic cleaning function (At this point, the dispatch center / actuator support automatic log cleanup, expired a few days a default setting is 30 days): New configuration items ( "xxl.job.logretentiondays") dispatch center logs stored for several days , expired logs automatically clean; to address the situation under massive log log table slow SQL problem; restrictions on entry into force of 7 or greater, or turn off cleaning function, the default is 30;
- 2, the report scheduling optimization: storing new table log report, the job log in three days at a frequency of once per minute synchronous to asynchronous reporting; Task Report read only report data, greatly enhance the speed of loading;
- 3, Cron online generation tools: Task add, edit box to generate online component Cron expression;
- 4, Cron next query execution time: 5 times in a row to support the implementation of the follow-up time to view online through the interface;
- 5, the new dispatch center application health checks, by means of "spring-boot-starter-actuator", the relative address "/ actuator / health";
- 6, DB script default encoding instead utf8mb4, repair garbled character issue (recommendation Mysql version 5.7+);
- 7, the central task scheduling evenly distributed, the number of trigger thread pool tasks related to the number of components and each acquisition, avoid large number of tasks in a single centralized dispatch center cluster node;
- 8, the optimization task trigger assembly, a preload 1s normal frequency, when a preload bye active sleep cycle loading, thus reducing dynamic loading frequency reduced pressure DB;
- 9, optimize the scheduling component: Prohibition to configure and start for Cron never triggered; Cron task is triggered after the last time it will not trigger, such as a one-time task, take the initiative to stop the related tasks;
- 10, DB reconnect optimize, repair, DB task scheduling problem stopped after reconnection downtime, automatic reconnect after adding trigger cluster scheduling task scheduling;
- 11, registration monitoring thread optimization, reducing the chance of deadlock;
- 12, delete the log dispatch center optimization, instead paging get the ID and ID to delete the manner, prevent bulk deletions cause massive log deadlock;
- 13, when the task retry parameter is missing bug fixes;
- 14, the control center is removed in SQL "now ()" function; no longer dependent on the cluster DB clock deployment is only necessary to ensure that the dispatch center to consistent application node clock;
- 15, task trigger adjustment assembly loading order to avoid a small probability that a random component to load I18N order NPE problem caused;
- 16, JobThread self-destruction optimization, to avoid triggering cause concurrent tasks triggerQueue loss problem;
- 17, 18 password restrictions dispatch center, repair to change the password over 18 could not land in question;
- 18, task alarms components paging parameters invalidity repair;
- 19, version upgrade xxl-rpc: server thread optimized to reduce memory overhead thread; IpUtil optimization: increase the connectivity of schools, filtered clear illegal card;
- 20, dispatch center callback API services instead restful way;
- 21, UI optimization, task lists and list data log table width scaling, to avoid data line feed to enhance the experience;
- 22, fill in the login screen to cancel the default login account and password;
- 23, actuators adjust the properties of the table, "order" attribute adjusted to integer, solve problems more actuators not sort the data correctly;
- 24, interactive task list optimization, support the view registration node task belongs actuator;
- 25, the project relies upgrade to a newer stable version, such as spring, spring-boot, mybatis, slf4j, groovy and so on;
Brief introduction
XXL-JOB is a lightweight distributed task scheduling platform, its core design goal is to develop rapid, simple to learn, lightweight, easy to expand. Open source and now online companies access to the product line, out of the box.
characteristic
- 1, simple: support for CRUD operations tasks through the Web page, simple operation, one minute to get started;
- 2, dynamic: support dynamic modification of the task status, start / stop tasks, and terminates the mandate with immediate effect;
- 3, the dispatch center HA (centric): dispatch center design using "control center" RESEARCH scheduling component from cluster deployment and support, the HA may be guaranteed scheduling center;
- 4, the actuator HA (distributed): distributed task execution, the task "actuator" supports cluster deployment, can guarantee task execution HA;
- 5, the registry: Actuator periodically automatically registered task, the dispatch center will be registered automatically find and trigger the execution of the task. It also supports manual entry actuators address;
- 6, the elastic expansion volume reduction: Once a new line or offline actuators on the machine, the task will reassign the next scheduling;
- 7, routing strategy: provide rich routing policy cluster deployment actuator comprising: first, last, polling, random, consistency HASH, most do not often use, least recently used failover, busy transfer, etc. ;
- 8, failover: Routing Policy Task select "failover" case, the actuator if one of the machines in the cluster fails, it will automatically switch to a normal Failover actuator transmits a scheduling request.
- 9, the blocking processing policy: when the scheduling processing strategy is too late to perform processing too dense, strategies include: single serial (default), discard subsequent scheduling, scheduling before overwriting;
- 10, the task timeout control: support custom task timeout, the task will take the initiative to run a timeout interrupt tasks;
- 11, the task failure retry: support custom failure retry task, when the task will fail according to a preset failure retry active retry; wherein the support slice task failed retry fragment size;
- 12, the mission fails alarm; the default mail failed to provide warning, while reserve expansion interface allows easy expansion of SMS, nails and other alarm mode;
- 13, fragmented broadcasting task: when a cluster deployment actuator, select the next task routing policy "fragmented broadcasting" situation, a task scheduler will trigger a cluster broadcast all actuators to perform a task, according to the fragment parameters developed slice task ;
- 14, the dynamic fragmentation: fragmentation broadcasting task to the actuator as a dimension fragmentation, supports dynamic expansion actuator cluster dynamically increasing the number of fragments, collaborative business processes; large volume of data traffic operation is performed to significantly improve tasking capacity and speed.
- 15, event triggers: In addition to "Cron way" and "task-dependent manner" trigger mandate, to support event-based trigger mode task. Dispatch centers provide services API to perform a single task is triggered, the trigger can be flexibly according to business events.
- 16, task progress monitoring: real-time monitoring of the progress of the task;
- 17, Rolling real-time log: support online viewing scheduling results, and support for real-time viewing mode Rolling actuator output complete execution log;
- 18, GLUE: providing Web IDE, support online development tasks logic code, dynamic publishing, real-time compiler into force, the deployment process is omitted on the line. It supports 30 versions of historical versions back.
- 19, Script task: supporting the development of GLUE mode and run a script task, including Shell, Python, NodeJS, PHP, PowerShell and other types of scripts;
- 20, the command line tasks: providing universal primary command-line task Handler (Bean task, "CommandJobHandler"); business only needs to provide command lines;
- 21, task dependencies: Support Configuration sub-task dependencies, when the parent task execution and after the end of the successful implementation of the initiative will trigger the execution of a sub-task, multiple sub-tasks, separated by commas;
- 22, consistency: "control center" to ensure the consistency of the cluster distributed scheduling by DB lock, a task scheduler triggers only one execution;
- 23, custom task parameters: task scheduling supports online configuration parameters into immediate effect;
- 24, the thread pool scheduling: scheduling system triggered scheduling multiple threads running, ensure accurate implementation of scheduling, not blocked;
- 25, data encryption: communication between the control center and the actuator for data encryption, scheduling information to enhance security;
- 26, e-mail alarm: Support mission failure alarm messages, support for configuring multiple e-mail addresses bulk alarm messages;
- 27, push maven central warehouse: the latest stable version will be pushed to the maven central repository, user-friendly access and use;
- 28, run the report: view real-time performance data, such as the number of tasks, scheduling times, number of actuators; and schedule reports, such as the dispatch date maps, maps and other scheduling success;
- 29, all asynchronous: task scheduling process to achieve full asynchronous design, such as asynchronous scheduling, asynchronous operation, such as an asynchronous callback, effective scheduling of intensive traffic clipping, in theory, support the long-running tasks arbitrary;
- 30, cross-platform: providing universal primary task HTTP Handler (Bean task, "HttpJobHandler"); business only needs to provide HTTP link, does not limit the language, platform;
- 31, international: the dispatch center to support international settings, providing Chinese and English language optional, defaults to Chinese;
- 32, of the container: official docker mirror, and updated in real time push dockerhub, further achieve the product out of the box;
- 33, the thread pool isolation: isolation scheduled thread pool split into the slow task automatically downgrade "Slow" thread pool to avoid depletion of the thread scheduling, to improve the stability of the system ;;
- 34, User Management: Support Online user management system, there are administrators, ordinary users two roles;
- 35, access control: access control actuator dimensions, the administrator has the full amount of authority, ordinary users need to be assigned permissions before allowing the actuator related operations;