Skip to content
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

fix(qrm): fix sharedCoresWithNUMABindingHintHandler when hints alread… #785

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix(qrm): fix sharedCoresWithNUMABindingHintHandler when hints alread… #785

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
60 changes: 30 additions & 30 deletions pkg/agent/qrm-plugins/cpu/dynamicpolicy/policy_hint_handlers.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -718,44 +718,44 @@ func (p *DynamicPolicy) sharedCoresWithNUMABindingHintHandler(_ context.Context,
req.PodNamespace, req.PodName, req.ContainerName, err)
return nil, fmt.Errorf("GenerateMachineStateFromPodEntries failed with error: %v", err)
}
}

general.Infof("pod: %s/%s, container: %s, inplace update resize: %v", req.PodNamespace, req.PodName, req.ContainerName, util.PodInplaceUpdateResizing(req))
numaSet := allocationInfo.GetAllocationResultNUMASet()
if numaSet.Size() != 1 {
general.Errorf("pod: %s/%s, container: %s is snb, but its numa set size is %d",
req.PodNamespace, req.PodName, req.ContainerName, numaSet.Size())
return nil, fmt.Errorf("snb port not support cross numa")
}
nodeID := numaSet.ToSliceInt()[0]
availableCPUQuantity := machineState[nodeID].GetAvailableCPUQuantity(p.reservedCPUs)
general.Infof("pod: %s/%s, container: %s, inplace update resize: %v", req.PodNamespace, req.PodName, req.ContainerName, util.PodInplaceUpdateResizing(req))
numaSet := allocationInfo.GetAllocationResultNUMASet()
if numaSet.Size() != 1 {
general.Errorf("pod: %s/%s, container: %s is snb, but its numa set size is %d",
req.PodNamespace, req.PodName, req.ContainerName, numaSet.Size())
return nil, fmt.Errorf("snb port not support cross numa")
}
nodeID := numaSet.ToSliceInt()[0]
availableCPUQuantity := machineState[nodeID].GetAvailableCPUQuantity(p.reservedCPUs)

general.Infof("pod: %s/%s, container: %s request cpu inplace update resize on numa %d (available: %.3f, request: %.3f)",
req.PodNamespace, req.PodName, req.ContainerName, nodeID, availableCPUQuantity, request)
if !cpuutil.CPUIsSufficient(request, availableCPUQuantity) { // no left resource to scale out
general.Infof("pod: %s/%s, container: %s request cpu inplace update resize, but no enough resource for it in current NUMA, checking migratable",
req.PodNamespace, req.PodName, req.ContainerName)
// TODO move this var to config
isInplaceUpdateResizeNumaMigration := false
if isInplaceUpdateResizeNumaMigration {
general.Infof("pod: %s/%s, container: %s request inplace update resize and no enough resource in current NUMA, try to migrate it to new NUMA",
general.Infof("pod: %s/%s, container: %s request cpu inplace update resize on numa %d (available: %.3f, request: %.3f)",
req.PodNamespace, req.PodName, req.ContainerName, nodeID, availableCPUQuantity, request)
if !cpuutil.CPUIsSufficient(request, availableCPUQuantity) { // no left resource to scale out
general.Infof("pod: %s/%s, container: %s request cpu inplace update resize, but no enough resource for it in current NUMA, checking migratable",
req.PodNamespace, req.PodName, req.ContainerName)
var calculateErr error
hints, calculateErr = p.calculateHintsForNUMABindingSharedCores(request, podEntries, machineState, req)
if calculateErr != nil {
general.Errorf("pod: %s/%s, container: %s request inplace update resize and no enough resource in current NUMA, failed to migrate it to new NUMA",
// TODO move this var to config
isInplaceUpdateResizeNumaMigration := false
if isInplaceUpdateResizeNumaMigration {
general.Infof("pod: %s/%s, container: %s request inplace update resize and no enough resource in current NUMA, try to migrate it to new NUMA",
req.PodNamespace, req.PodName, req.ContainerName)
return nil, fmt.Errorf("calculateHintsForNUMABindingSharedCores failed in inplace update resize mode with error: %v", calculateErr)
var calculateErr error
hints, calculateErr = p.calculateHintsForNUMABindingSharedCores(request, podEntries, machineState, req)
if calculateErr != nil {
general.Errorf("pod: %s/%s, container: %s request inplace update resize and no enough resource in current NUMA, failed to migrate it to new NUMA",
req.PodNamespace, req.PodName, req.ContainerName)
return nil, fmt.Errorf("calculateHintsForNUMABindingSharedCores failed in inplace update resize mode with error: %v", calculateErr)
}
} else {
general.Errorf("pod: %s/%s, container: %s request inplace update resize, but no enough resource for it in current NUMA",
req.PodNamespace, req.PodName, req.ContainerName)
return nil, errNoAvailableCPUHints
}
} else {
general.Errorf("pod: %s/%s, container: %s request inplace update resize, but no enough resource for it in current NUMA",
general.Infof("pod: %s/%s, container: %s request inplace update resize, there is enough resource for it in current NUMA",
req.PodNamespace, req.PodName, req.ContainerName)
return nil, errNoAvailableCPUHints
hints = cpuutil.RegenerateHints(allocationInfo, false)
}
} else {
general.Infof("pod: %s/%s, container: %s request inplace update resize, there is enough resource for it in current NUMA",
req.PodNamespace, req.PodName, req.ContainerName)
hints = cpuutil.RegenerateHints(allocationInfo, false)
}
} else {
var calculateErr error
Expand Down
Loading