Skip to content

Consider all machines of partition for evaluating faulty machines #632

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Gerrit91 merged 4 commits into from
Sep 22, 2025
Merged

Consider all machines of partition for evaluating faulty machines #632

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
242f99b
consider all machines of partition for evaluating faulty machiens
ostempel Sep 2, 2025
e440594
Merge branch 'master' of github.com:metal-stack/metal-api into fix-ca…
ostempel Sep 2, 2025
0634c22
fix all machines query
ostempel Sep 3, 2025
621e7ca
Merge branch 'master' into fix-capacity-endpoint-faulty-machine-evalu…
Gerrit91 Sep 22, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 14 additions & 3 deletions cmd/metal-api/internal/service/partition-service.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -380,8 +380,9 @@ func (r *partitionResource) partitionCapacity(request *restful.Request, response

func (r *partitionResource) calcPartitionCapacity(pcr *v1.PartitionCapacityRequest) ([]v1.PartitionCapacity, error) {
var (
ps metal.Partitions
ms metal.Machines
ps metal.Partitions
ms metal.Machines
allMs metal.Machines

pcs = map[string]*v1.PartitionCapacity{}

Expand Down Expand Up @@ -413,6 +414,16 @@ func (r *partitionResource) calcPartitionCapacity(pcr *v1.PartitionCapacityReque
return nil, err
}

// if filtered on partition get all without more filters for issues evaluation
if machineQuery.PartitionID != nil {
err = r.ds.SearchMachines(&datastore.MachineSearchQuery{PartitionID: machineQuery.PartitionID}, &allMs)
if err != nil {
return nil, err
}
} else {
allMs = ms
Copy link
Contributor

@Gerrit91 Gerrit91 Sep 2, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

ms could be filtered by size, so these are not all machines. There should be no other filters applied for machine evaluation (except maybe partition).

Copy link
Contributor Author

@ostempel ostempel Sep 2, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yeah totally right. I will fix this

}

ecs, err := r.ds.ListProvisioningEventContainers()
if err != nil {
return nil, fmt.Errorf("unable to fetch provisioning event containers: %w", err)
Expand All @@ -429,7 +440,7 @@ func (r *partitionResource) calcPartitionCapacity(pcr *v1.PartitionCapacityReque
}

machinesWithIssues, err := issues.Find(&issues.Config{
Machines: ms,
Machines: allMs,
EventContainers: ecs,
Omit: []issues.Type{issues.TypeLastEventError},
})
Expand Down
107 changes: 107 additions & 0 deletions cmd/metal-api/internal/service/partition-service_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -310,6 +310,27 @@ func TestPartitionCapacity(t *testing.T) {
mock.On(r.DB("mockdb").Table("event")).Return(events, nil)
mock.On(r.DB("mockdb").Table("partition")).Return(partitions, nil)
mock.On(r.DB("mockdb").Table("size")).Return(sizes, nil)

//Filtered mocks
var filteredPartition []metal.Partition
for _, partition := range partitions {
if partition.ID == "partition-a" {
filteredPartition = append(filteredPartition, partition)
}
}
var filteredMachinesByPartition []metal.Machine
var filteredMachinesByPartitionBySize []metal.Machine
for _, machine := range ms {
if machine.PartitionID == "partition-a" {
filteredMachinesByPartition = append(filteredMachinesByPartition, machine)
if machine.SizeID == "size-a" {
filteredMachinesByPartitionBySize = append(filteredMachinesByPartitionBySize, machine)
}
}
}
mock.On(r.DB("mockdb").Table("partition").Get("partition-a")).Return(filteredPartition, nil)
mock.On(r.DB("mockdb").Table("machine").Filter(func(var_1 r.Term) r.Term { return var_1.Field("partitionid").Eq("partition-a") })).Return(filteredMachinesByPartition, nil)
mock.On(r.DB("mockdb").Table("machine").Filter(func(var_1 r.Term) r.Term { return var_1.Field("partitionid").Eq("partition-a") }).Filter(func(var_1 r.Term) r.Term { return var_1.Field("sizeid").Eq("size-a") })).Return(filteredMachinesByPartitionBySize, nil)
}

machineTpl = func(id, partition, size, project string) metal.Machine {
Expand Down Expand Up @@ -411,6 +432,92 @@ func TestPartitionCapacity(t *testing.T) {
},
},
},
{
name: "filter considers all machines",
pcr: &v1.PartitionCapacityRequest{
ID: pointer.Pointer("partition-a"),
Size: pointer.Pointer("size-a"),
},
mockFn: func(mock *r.Mock) {
m1 := machineTpl("1", "partition-a", "size-a", "project-123")
m2 := machineTpl("2", "partition-a", "size-a", "project-123")
m3 := machineTpl("3", "partition-a", "size-a", "project-123")
m4 := machineTpl("4", "partition-a", "size-b", "project-123")
m4.IPMI.Address = "1.2.3.1"
m5 := machineTpl("5", "partition-b", "size-a", "project-123")
mockMachines(mock, metal.MachineLivelinessAlive, nil, m1, m2, m3, m4, m5)
},

want: []*v1.PartitionCapacity{
{
Common: v1.Common{
Identifiable: v1.Identifiable{ID: "partition-a"}, Describable: v1.Describable{Name: pointer.Pointer(""), Description: pointer.Pointer("")},
},
ServerCapacities: v1.ServerCapacities{
{
Size: "size-a",
Total: 3,
PhonedHome: 3,
Faulty: 1,
Allocated: 3,
FaultyMachines: []string{"1"},
},
},
},
},
},
{
name: "non filter considers all machines",
pcr: &v1.PartitionCapacityRequest{},
mockFn: func(mock *r.Mock) {
m1 := machineTpl("1", "partition-a", "size-a", "project-123")
m2 := machineTpl("2", "partition-a", "size-a", "project-123")
m3 := machineTpl("3", "partition-a", "size-a", "project-123")
m4 := machineTpl("4", "partition-a", "size-b", "project-123")
m4.IPMI.Address = "1.2.3.1"
m5 := machineTpl("5", "partition-b", "size-a", "project-123")
mockMachines(mock, metal.MachineLivelinessAlive, nil, m1, m2, m3, m4, m5)
},
want: []*v1.PartitionCapacity{
{
Common: v1.Common{
Identifiable: v1.Identifiable{ID: "partition-a"}, Describable: v1.Describable{Name: pointer.Pointer(""), Description: pointer.Pointer("")},
},
ServerCapacities: v1.ServerCapacities{
{
Size: "size-a",
Total: 3,
PhonedHome: 3,
Faulty: 1,
Allocated: 3,
FaultyMachines: []string{"1"},
},
{
Size: "size-b",
Total: 1,
PhonedHome: 1,
Faulty: 1,
Allocated: 1,
FaultyMachines: []string{"4"},
},
},
},
{
Common: v1.Common{
Identifiable: v1.Identifiable{ID: "partition-b"}, Describable: v1.Describable{Name: pointer.Pointer(""), Description: pointer.Pointer("")},
},
ServerCapacities: v1.ServerCapacities{
{
Size: "size-a",
Total: 1,
PhonedHome: 1,
Faulty: 0,
Allocated: 1,
},
},
},
},
},
{
name: "one waiting machine",
mockFn: func(mock *r.Mock) {
Expand Down