Thanks to visit codestin.com
Credit goes to github.com

Skip to content

[release-1.21] Bug 1977100: retry setting CPU load balancing #5919

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

Closed
openshift-cherrypick-robot wants to merge 1 commit into from
+23 −3
Closed

[release-1.21] Bug 1977100: retry setting CPU load balancing #5919

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
26 changes: 23 additions & 3 deletions internal/runtimehandlerhooks/high_performance_hooks.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@ import (
"path/filepath"
"strconv"
"strings"
"time"

"github.com/cri-o/cri-o/internal/config/cgmgr"
"github.com/cri-o/cri-o/internal/lib/sandbox"
Expand All @@ -21,6 +22,7 @@ import (
"github.com/pkg/errors"
"github.com/sirupsen/logrus"
"k8s.io/apimachinery/pkg/fields"
"k8s.io/apimachinery/pkg/util/wait"
"k8s.io/kubernetes/pkg/kubelet/cm/cpuset"
)

Expand Down Expand Up @@ -65,8 +67,7 @@ func (h *HighPerformanceHooks) PreStart(ctx context.Context, c *oci.Container, s

// disable the CPU load balancing for the container CPUs
if shouldCPULoadBalancingBeDisabled(s.Annotations()) {
log.Infof(ctx, "Disable cpu load balancing for container %q", c.ID())
if err := setCPUSLoadBalancing(c, false, schedDomainDir); err != nil {
if err := setCPUSLoadBalancingWithRetry(ctx, c, false); err != nil {
return errors.Wrap(err, "set CPU load balancing")
}
}
Expand Down Expand Up @@ -112,7 +113,7 @@ func (h *HighPerformanceHooks) PreStop(ctx context.Context, c *oci.Container, s

// enable the CPU load balancing for the container CPUs
if shouldCPULoadBalancingBeDisabled(s.Annotations()) {
if err := setCPUSLoadBalancing(c, true, schedDomainDir); err != nil {
if err := setCPUSLoadBalancingWithRetry(ctx, c, true); err != nil {
return errors.Wrap(err, "set CPU load balancing")
}
}
Expand Down Expand Up @@ -172,6 +173,25 @@ func isContainerRequestWholeCPU(c *oci.Container) bool {
return *(c.Spec().Linux.Resources.CPU.Shares)%1024 == 0
}

func setCPUSLoadBalancingWithRetry(ctx context.Context, c *oci.Container, enable bool) error {
log.Infof(ctx, "Disable cpu load balancing for container %q", c.ID())
// it is possible to have errors during reading or writing to sched_domain files because
// that kernel rebuilds it with updated values
// the retry will not fix it for 100% but should reduce the possibility for failures to minimum
// TODO: re-visit once we will have some more acceptable cgroups hierarchy to disable CPU load balancing
// correctly via cgroups, see -https://bugzilla.redhat.com/show_bug.cgi?id=1946801
return wait.PollImmediate(time.Second, 5*time.Second, func() (bool, error) {
if err := setCPUSLoadBalancing(c, enable, schedDomainDir); err != nil {
if os.IsNotExist(err) {
log.Errorf(ctx, "Failed to set CPU load balancing: %v", err)
return false, nil
}
return false, err
}
return true, nil
})
}

func setCPUSLoadBalancing(c *oci.Container, enable bool, schedDomainDir string) error {
lspec := c.Spec().Linux
if lspec == nil ||
Expand Down