Serial scheduling

[weekface]

This pr fixes: #264, make sure only one pod can be scheduled at the same time.

[weekface]

/run-e2e-tests

[weekface]

/run-e2e-tests

[weekface]

/run-e2e-tests

[onlymellb]

Why do you want to add the condition currentPVC == nil, will there be multiple PVCs with the same name?

[weekface]

Removed.

[onlymellb]

Suggested change

	for i := 0; i < len(pvcList.Items); i++ {
	for _, pvc := range pvcList.Items {

[weekface]

Use this style to avoid the golang common mistakes

[onlymellb]

There is not use closures running as goroutines，If there is such a scene, then the way to use range can also avoid this problem.

[weekface]

currentPVC and schedulingPVC are pointers, the pvc in the for range block is a variable, if we use the for range style, then these two variables will get the same address.

[weekface]

For example:

package main

import "fmt"

func main() {
	arr := []string{"a", "b", "c"}

	var pa *string
	var pb *string
	for _, item := range arr {
		if item == "a" {
			pa = &item
		}
		if item == "b" {
			pb = &item
		}
	}
	fmt.Println(pa)
	fmt.Println(pb)
	fmt.Println(pa == pb)

	var ppa *string
	var ppb *string
	for i := 0; i < len(arr); i++ {
		if arr[i] == "a" {
			ppa = &arr[i]
		}
		if arr[i] == "b" {
			ppb = &arr[i]
		}
	}
	fmt.Println(ppa)
	fmt.Println(ppb)
	fmt.Println(ppa == ppb)
}

Outputs:

0xc00005e1c0
0xc00005e1c0
true
0xc000060180
0xc000060190
false

[onlymellb]

You can create a new variable to fix this problem. such as follow:

for i := 0; i < len(pvcList.Items); i++ { 
    pvc := pvcList.Items[i]
    ...
}
------
for _, pvc := range pvcList.Items {
    pvc := pvc
 ...
}

[weekface]

Style for i:=0; i<n; i++ is equivalent to the for range style, but using for i:=0; i<n; i++ is more clear at this situation.

pvc := pvc looks weird.

[xiaojingchen]

LGTM

[onlymellb]

LGTM

[tennix]

LGTM

[tennix]

/run-e2e-tests

[weekface]

/run-e2e-tests