README
¶
rust version: weighted-rs
Package weighted implements the smooth weighted round-robin balancing algorithm. This algorithm is implemented in Nginx: https://github.com/phusion/nginx/commit/27e94984486058d73157038f7950a0a36ecc6e35.
Notice: The weighted is NOT goroutine-safe so you MUST use the synchronization primitive to protect it (the Next method) in concurrent cases.
Algorithm is as follows: on each peer selection we increase current_weight of each eligible peer by its weight, select peer with greatest current_weight and reduce its current_weight by total number of weight points distributed among peers.
In case of { 5, 1, 1 } weights this gives the following sequence of current_weight's: (a, a, b, a, c, a, a)
This is an example to use it:
package main
import "fmt"
func ExampleW1_Next() {
w := &SW{}
w.Add("a", 5)
w.Add("b", 2)
w.Add("c", 3)
for i := 0; i < 10; i++ {
fmt.Printf("%s ", w.Next())
}
}
And this lib has provides another weighted round robin algorithm. This algorithm is used in LVS. It has better performance but it is not so more smooth than the first algorithm, so you can select one algorithm according to your case. It is used like the first:
package main
import "fmt"
func ExampleW2_Next() {
w := &W2{}
w.Add("a", 5)
w.Add("b", 2)
w.Add("c", 3)
for i := 0; i < 10; i++ {
fmt.Printf("%s ", w.Next())
}
}
Documentation
¶
Overview ¶
Package weighted implements two weighted round robin algorithms. One is the smooth weighted round-robin balancing algorithm used in Nginx, and you can use "w := new SW{}" to use it. The other is wrr used in LVS and you can use "w := new RRW{}" to use it.
For Nginx smooth weighted round-robin balancing algorithm, you can check https://github.com/phusion/nginx/commit/27e94984486058d73157038f7950a0a36ecc6e35. For LVS round-robin balancing algorithm, you can check http://kb.linuxvirtualitem.org/wiki/Weighted_Round-Robin_Scheduling.
Index ¶
Examples ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type RRW ¶
type RRW struct {
// contains filtered or unexported fields
}
RRW is a struct that contains weighted items implement LVS weighted round robin algorithm.
http://kb.linuxvirtualitem.org/wiki/Weighted_Round-Robin_Scheduling http://zh.linuxvirtualitem.org/node/37
type RandW ¶
type RandW struct {
// contains filtered or unexported fields
}
RandW is a struct that contains weighted items implement weighted random algorithm.
type SW ¶
type SW struct {
Count int
// contains filtered or unexported fields
}
SW (Smooth Weighted) is a struct that contains weighted items and provides methods to select a weighted item. It is used for the smooth weighted round-robin balancing algorithm. This algorithm is implemented in Nginx: https://github.com/phusion/nginx/commit/27e94984486058d73157038f7950a0a36ecc6e35.
Algorithm is as follows: on each peer selection we increase current_weight of each eligible peer by its weight, select peer with greatest current_weight and reduce its current_weight by total number of weight points distributed among peers.
In case of { 5, 1, 1 } weights this gives the following sequence of current_weight's: (a, a, b, a, c, a, a)
type W ¶
type W interface {
// Next gets next selected item.
// Next is not goroutine-safe. You MUST use the snchronization primitive to protect it in concurrent cases.
Next() (item interface{})
// Add adds a weighted item for selection.
Add(item interface{}, weight int)
// All returns all items.
All() map[interface{}]int
// RemoveAll removes all weighted items.
RemoveAll()
// Reset resets the balancing algorithm.
Reset()
}
W is a interface that implement a weighted round robin algorithm.
Source Files