背景

分布式任务分配

在很多运维场景下，我们都会执行一些长时间的任务，比如装机、部署环境、打包镜像等长时间任务， 而通常我们的任务节点数量通常是有限的(排除基于k8s的hpa、或者knative等自动伸缩场景)。

那么当我们有一个任务如何根据当前的worker和corrdinator和任务来进行合理的分配，分配其实也比较复杂，往复杂里面做，可以根据当前系统的负载、每个任务的执行资源消耗、当前集群的任务数量等， 这里我们就搞一个最简单的，基于任务和当前worker的RR算法

系统架构

在worker和任务队列之间，添加一层协调调度层Coordinator， 由它来根据当前集群任务的状态来进行任务的分配，同时感知当前集群worker和task的状态，协调整个集群任务的执行、终止等操作

单机实现

整体设计

members: 表示当前集群中所有的worker tasks: 就是当前的任务 Coordinator: 就是我们的协调者， 负责根据members和tasks进行任务的分配 result: 就是分配的结果

CircularIterator

CircularIterator就是我们的环状对立迭代器, 拥有两个方法， 一个是add添加member, 一个Next返回基于rr的下一个member

// CircularIterator 环状迭代器 type CircularIterator struct { list []interface{} // 保存所有的成员变量 next int } // Next 返回下一个元素 func (c *CircularIterator) Next() interface{} { item := c.list[c.next] c.next = (c.next + 1) % len(c.list) return item } // Add 添加任务 func (c *CircularIterator) Add(v interface{}) bool { for _, item := range c.list { if v == item { return false } } c.list = append(c.list, v) return true } 

Member&Task

Member就是负责执行任务的worker, 有一个AddTask方法和Execute方法负责任务的执行和添加任务 Task标识一个任务

 // Member 任务组成员 type Member struct { id int tasks []*Task } // ID 返回当前memberID func (m *Member) ID() int { return m.id } // AddTask 为member添加任务 func (m *Member) AddTask(t *Task) bool { for _, task := range m.tasks { if task == t { return false } } m.tasks = append(m.tasks, t) return true } // Execute 执行任务 func (m *Member) Execute() { for _, task := range m.tasks { fmt.Printf("Member %d run task %s\n", m.ID(), task.Execute()) } } // Task 任务 type Task struct { name string } // Execute 执行task返回结果 func (t *Task) Execute() string { return "Task " + t.name + " run success" } 

Coordinator

Coordinator是协调器，负责根据 Member和task进行集群任务的协调调度

// Task 任务 type Task struct { name string } // Execute 执行task返回结果 func (t *Task) Execute() string { return "Task " + t.name + " run success" } // Coordinator 协调者 type Coordinator struct { members []*Member tasks []*Task } // TaskAssignments 为member分配任务 func (c *Coordinator) TaskAssignments() map[int]*Member { taskAssignments := make(map[int]*Member) // 构建迭代器 memberIt := c.getMemberIterator() for _, task := range c.tasks { member := memberIt.Next().(*Member) _, err := taskAssignments[member.ID()] if err == false { taskAssignments[member.ID()] = member } member.AddTask(task) } return taskAssignments } func (c *Coordinator) getMemberIterator() *CircularIterator { // 通过当前成员, 构造成员队列 members := make([]interface{}, len(c.members)) for index, member := range c.members { members[index] = member } return NewCircularIterftor(members) } // AddMember 添加member组成员 func (c *Coordinator) AddMember(m *Member) bool { for _, member := range c.members { if member == m { return false } } c.members = append(c.members, m) return true } // AddTask 添加任务 func (c *Coordinator) AddTask(t *Task) bool { for _, task := range c.tasks { if task == t { return false } } c.tasks = append(c.tasks, t) return true } 

测试

我们首先创建一堆member和task, 然后调用coordinator进行任务分配，执行任务结果

 coordinator := NewCoordinator() for i := 0; i < 10; i++ { m := &Member{id: i} coordinator.AddMember(m) } for i := 0; i < 30; i++ { t := &Task{name: fmt.Sprintf("task %d", i)} coordinator.AddTask(t) } result := coordinator.TaskAssignments() for _, member := range result { member.Execute() } 

结果

可以看到每个worker均匀的得到任务分配

Member 6 run task Task task 6 run success Member 6 run task Task task 16 run success Member 6 run task Task task 26 run success Member 8 run task Task task 8 run success Member 8 run task Task task 18 run success Member 8 run task Task task 28 run success Member 0 run task Task task 0 run success Member 0 run task Task task 10 run success Member 0 run task Task task 20 run success Member 3 run task Task task 3 run success Member 3 run task Task task 13 run success Member 3 run task Task task 23 run success Member 4 run task Task task 4 run success Member 4 run task Task task 14 run success Member 4 run task Task task 24 run success Member 7 run task Task task 7 run success Member 7 run task Task task 17 run success Member 7 run task Task task 27 run success Member 9 run task Task task 9 run success Member 9 run task Task task 19 run success Member 9 run task Task task 29 run success Member 1 run task Task task 1 run success Member 1 run task Task task 11 run success Member 1 run task Task task 21 run success Member 2 run task Task task 2 run success Member 2 run task Task task 12 run success Member 2 run task Task task 22 run success Member 5 run task Task task 5 run success Member 5 run task Task task 15 run success Member 5 run task Task task 25 run success 

完整代码

package main import "fmt" // CircularIterator 环状迭代器 type CircularIterator struct { list []interface{} next int } // Next 返回下一个元素 func (c *CircularIterator) Next() interface{} { item := c.list[c.next] c.next = (c.next + 1) % len(c.list) return item } // Add 添加任务 func (c *CircularIterator) Add(v interface{}) bool { for _, item := range c.list { if v == item { return false } } c.list = append(c.list, v) return true } // Member 任务组成员 type Member struct { id int tasks []*Task } // ID 返回当前memberID func (m *Member) ID() int { return m.id } // AddTask 为member添加任务 func (m *Member) AddTask(t *Task) bool { for _, task := range m.tasks { if task == t { return false } } m.tasks = append(m.tasks, t) return true } // Execute 执行任务 func (m *Member) Execute() { for _, task := range m.tasks { fmt.Printf("Member %d run task %s\n", m.ID(), task.Execute()) } } // Task 任务 type Task struct { name string } // Execute 执行task返回结果 func (t *Task) Execute() string { return "Task " + t.name + " run success" } // Coordinator 协调者 type Coordinator struct { members []*Member tasks []*Task } // TaskAssignments 为member分配任务 func (c *Coordinator) TaskAssignments() map[int]*Member { taskAssignments := make(map[int]*Member) // 构建迭代器 memberIt := c.getMemberIterator() for _, task := range c.tasks { member := memberIt.Next().(*Member) _, err := taskAssignments[member.ID()] if err == false { taskAssignments[member.ID()] = member } member.AddTask(task) } return taskAssignments } func (c *Coordinator) getMemberIterator() *CircularIterator { // 通过当前成员, 构造成员队列 members := make([]interface{}, len(c.members)) for index, member := range c.members { members[index] = member } return NewCircularIterftor(members) } // AddMember 添加member组成员 func (c *Coordinator) AddMember(m *Member) bool { for _, member := range c.members { if member == m { return false } } c.members = append(c.members, m) return true } // AddTask 添加任务 func (c *Coordinator) AddTask(t *Task) bool { for _, task := range c.tasks { if task == t { return false } } c.tasks = append(c.tasks, t) return true } // NewCircularIterftor 返回迭代器 func NewCircularIterftor(list []interface{}) *CircularIterator { iterator := CircularIterator{} for _, item := range list { iterator.Add(item) } return &iterator } // NewCoordinator 返回协调器 func NewCoordinator() *Coordinator { c := Coordinator{} return &c } func main() { coordinator := NewCoordinator() for i := 0; i < 10; i++ { m := &Member{id: i} coordinator.AddMember(m) } for i := 0; i < 30; i++ { t := &Task{name: fmt.Sprintf("task %d", i)} coordinator.AddTask(t) } result := coordinator.TaskAssignments() for _, member := range result { member.Execute() } } 

总结

任务协调是一个非常复杂的事情， 内部的任务平台，虽然实现了基于任务的组合和app化，但是任务调度分配着一块，仍然没有去做，只是简单的根据树形任务去简单的做一些分支任务的执行，未来有时间再做吧，要继续研究下一个模块了

这个调度思想来源于kafka connect的DistributedHerder里面的WorkerCoordinator，感兴趣的可以看看，未完待续

更多文章可以访问http://www.sreguide.com/