Sysched: add the implementation of SySched #568

salmanyam · 2023-03-30T16:17:50Z

What type of PR is this?

/kind feature

What this PR does / why we need it:

Implementation of SySched, a system call-aware scheduler. The KEP of this PR is KEP #399.

Which issue(s) this PR fixes:

Fixes #

Special notes for your reviewer:

Thank you very much for reviewing this PR.
Please let us know if there are questions or comments.
Your comments and feedback are appreciated.

Does this PR introduce a user-facing change?

Introduce new Scoring plugin called SySched that will rank feasible nodes based on the relative risks of pods' system call usage.

linux-foundation-easycla · 2023-03-30T16:17:56Z

The committers listed above are authorized under a signed CLA.

✅ login: salmanyam / name: Salman Ahmed (1295000, 4b6ea59, 588b8ec, 41473cb)

k8s-ci-robot · 2023-03-30T16:18:00Z

Hi @salmanyam. Thanks for your PR.

I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

salmanyam · 2023-03-30T16:26:29Z

@Huang-Wei This is the implementation of KEP #399, a system call-aware scheduler plugin called SySched. Could you please take a look to see if there are any concerns or comments? Thank you!

Huang-Wei · 2023-04-05T19:13:04Z

Thanks for the PR. I'm a bit swamped recently. So if any reviewers have free cycles, please assign to yourself.

/help

PiotrProkop

My first pass on this PR. Didn't dive deep into the logic.

PiotrProkop · 2023-04-11T08:22:15Z

manifests/sysched/examples/memcached1.yaml

+      #serviceAccountName: default
+      #schedulerName: scheduler-plugins-scheduler


PiotrProkop · 2023-04-11T08:22:31Z

manifests/sysched/examples/nginx1.yaml

+      #serviceAccountName: default
+      #schedulerName: scheduler-plugins-scheduler


PiotrProkop · 2023-04-11T08:22:41Z

manifests/sysched/examples/nginx2.yaml

+      #serviceAccountName: default
+      #schedulerName: scheduler-plugins-scheduler


PiotrProkop · 2023-04-11T08:22:55Z

manifests/sysched/examples/redis1.yaml

+      #serviceAccountName: default
+      #schedulerName: scheduler-plugins-scheduler


PiotrProkop · 2023-04-11T08:23:01Z

manifests/sysched/examples/redis1.yaml

+        #resources:
+        #  requests:
+        #    cpu: 100m
+        #    memory: 100Mi


PiotrProkop · 2023-04-11T08:36:29Z

pkg/sysched/sysched.go

+	W := 1
+	score := len(syscalls) - tot_crit
+	score = score + W*tot_crit
+	klog.Info(score, " ", tot_crit)


I think this function will always return same score. tot_crit is always 0

Currently, our plugin does not support syscall weights. We are working on it and update the weighting in our next version. We leave that portion of code intentionally. Also, we have put a note there regarding it.

please add a comment explaining that weight is hard-coded for now.

PiotrProkop · 2023-04-11T08:37:29Z

pkg/sysched/sysched.go

+
+	nodeIP := node.Status.Addresses[0].Address
+
+	podSyscalls, _ := sc.getSyscalls(pod)


always check errors

PiotrProkop · 2023-04-11T08:38:15Z

pkg/sysched/sysched.go

+}
+
+func (sc *SySched) updateHostSyscalls(pod *v1.Pod) {
+	syscall, _ := sc.getSyscalls(pod)


always check errors

PiotrProkop · 2023-04-11T08:38:29Z

pkg/sysched/sysched.go

+	syscalls := make(map[string]bool)
+
+	for _, p := range pods {
+		syscall, _ := sc.getSyscalls(p)


always check errors

PiotrProkop · 2023-04-11T08:38:37Z

pkg/sysched/sysched.go

+		if p.Name == pod.Name {
+			sc.HostToPods[nodeIP] = remove(sc.HostToPods[nodeIP], i)
+			sc.HostSyscalls[nodeIP] = sc.recomputeHostSyscalls(sc.HostToPods[nodeIP])
+			c, _ := sc.getHostSyscalls(nodeIP)


always check errors

Huang-Wei · 2023-04-21T02:46:48Z

/ok-to-test

salmanyam · 2023-04-21T15:53:22Z

Thank you very much, @PiotrProkop! We have addressed all the comments. Please let us know if you have more comments or concerns.

salmanyam · 2023-05-22T15:11:24Z

Hi @Huang-Wei, we have rebased our code as the bot triggered need-rebase. Please let us know if there's anything needed from our end. Thanks!

Huang-Wei · 2023-09-18T18:01:45Z

pkg/sysched/sysched.go

+
+// Score invoked at the score extension point.
+func (sc *SySched) Score(ctx context.Context, cs *framework.CycleState, pod *v1.Pod, nodeName string) (int64, *framework.Status) {
+	node, err := sc.handle.ClientSet().CoreV1().Nodes().Get(context.TODO(), nodeName, metav1.GetOptions{})


You should be able to get cached NodeInfo from scheduler framework, which is the source of truth to do scheduling. (if reading from API server, the "cached" info would not be there)

nodeInfo, err := sc.handle.SnapshotSharedLister().NodeInfos().Get(nodeName)

We noticed the cached state is not always up-to-date, especially when there are activities on other schedulers in the system or when our scheduler comes up after a bunch of pods and nodes have already appeared, hence, we opted not to use the SnapshotSharedLister.

We tried inserting some comments in the code to explain our rationale. Please let us know if this makes sense.

Huang-Wei · 2023-09-18T18:03:24Z

pkg/sysched/sysched.go

+		return 0, nil
+	}
+
+	nodeIP := node.Status.Addresses[0].Address


Why using a node IP instead of node.Name as the canonical identifier to refer to a Node?

Huang-Wei · 2023-09-18T18:05:48Z

pkg/sysched/sysched.go

+type SySched struct {
+	handle                  framework.Handle
+	clientSet               v1alpha1.SPOV1Alpha1Interface
+	HostToPods              map[string][]*v1.Pod


the frameworkHandle already holds this: sc.handle.SnapshotSharedLister().

As explained above, we keep a local map of HostToPods in our plugin rather than rely on the cached state to reliably keep track of all pods on the cluster, regardless of which scheduler it used.

Huang-Wei · 2023-09-18T18:09:54Z

pkg/sysched/sysched.go

+	handle                  framework.Handle
+	clientSet               v1alpha1.SPOV1Alpha1Interface
+	HostToPods              map[string][]*v1.Pod
+	HostSyscalls            map[string]map[string]bool


could you add a comment explaining what's keyed, and what's valued?

Commented in another thread, in Golang we don't use map[string]bool to represent a set; instead, we use map[string]struct{}. And you even don't need to implement you self - reuse Set as well as its Untion/Intersect functions.

Huang-Wei · 2023-09-18T18:10:04Z

pkg/sysched/sysched.go

+	clientSet               v1alpha1.SPOV1Alpha1Interface
+	HostToPods              map[string][]*v1.Pod
+	HostSyscalls            map[string]map[string]bool
+	CritSyscalls            map[string][]string


Also comment it.

Huang-Wei · 2023-09-18T19:05:13Z

test/integration/sysched_test.go

+	}
+	cfg.Profiles[0].Plugins.Score = schedapi.PluginSet{
+		Enabled: []schedapi.Plugin{{Name: sysched.Name}},
+		//Disabled: []schedapi.Plugin{{Name: "*"}},


You should be disabling it. Otherwise score of other plugins may impact the eventual result.

Huang-Wei · 2023-09-18T19:05:39Z

test/integration/sysched_test.go

+	_, err = SPOCreate(extClient, &fullseccompSPOCR, metav1.CreateOptions{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	_, err = SPOCreate(extClient, &badSPOCR, metav1.CreateOptions{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	_, err = SPOCreate(extClient, &good1SPOCR, metav1.CreateOptions{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	_, err = SPOCreate(extClient, &good2SPOCR, metav1.CreateOptions{})
+	if err != nil {
+		t.Fatal(err)
+	}


De-duple the code by using a for loop.

Huang-Wei · 2023-09-18T19:08:16Z

pkg/sysched/sysched_test.go

could you make the test more table-driven? you can refer to other unit tests like pkg/capacityscheduling/capacityscheduling_test.go

Huang-Wei · 2023-09-18T19:09:06Z

pkg/sysched/sysched_test.go

+package sysched
+
+import (
+	//"fmt"


remove useless comments please (applies elsewhere).

Huang-Wei · 2023-09-18T19:10:14Z

pkg/sysched/sysched_test.go

+	v1 "k8s.io/api/core/v1"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/runtime/schema"
+	informers "k8s.io/client-go/informers"
+	clientsetfake "k8s.io/client-go/kubernetes/fake"
+	"k8s.io/client-go/kubernetes/scheme"
+	restclient "k8s.io/client-go/rest"
+	restfake "k8s.io/client-go/rest/fake"
+	"k8s.io/kubernetes/pkg/scheduler/framework"
+	"k8s.io/kubernetes/pkg/scheduler/framework/plugins/defaultbinder"
+	"k8s.io/kubernetes/pkg/scheduler/framework/plugins/queuesort"
+	frameworkruntime "k8s.io/kubernetes/pkg/scheduler/framework/runtime"
+	st "k8s.io/kubernetes/pkg/scheduler/testing"
+	"net/http"
+	pluginconfig "sigs.k8s.io/scheduler-plugins/apis/config"
+	v1alpha1 "sigs.k8s.io/scheduler-plugins/pkg/sysched/clientset/v1alpha1"
+	"sigs.k8s.io/security-profiles-operator/api/seccompprofile/v1beta1"
+	"strings"
+	"testing"


Please sort the imports by groups: (applies elsewhere)

native imports (followed an empty line)

upstream imports (followed by an empty line)

imports in this repo

mvle · 2023-09-21T19:30:59Z

Thanks @Huang-Wei for taking the time to review our patch. We should be able to address all of your concerns in the next PR.

netlify · 2023-10-24T15:39:37Z

✅ Deploy Preview for kubernetes-sigs-scheduler-plugins canceled.

Name	Link
🔨 Latest commit	`4b6ea59`
🔍 Latest deploy log	https://app.netlify.com/sites/kubernetes-sigs-scheduler-plugins/deploys/65e78e81366eb800087eefb3

mvle · 2023-11-02T01:44:15Z

@Huang-Wei Thanks for all your comments so far. We've updated the code to cover most of the comments, with some exceptions as explained in the comments above. Main items we have not changed:

SnapshotSharedLister for getting nodeInfo and HostToPods (opted not to use for reasons stated above).
custom clientset (WIP. Might need a PR to update SPO's code).

Please let us know your thoughts on the changes.

mvle · 2024-01-17T21:16:06Z

@Huang-Wei Hi! Wondering if you'll have a chance to take a look at this PR in the coming days? Looking forward to see if we can get this closer to being merged.

Huang-Wei · 2024-01-18T00:02:37Z

@mvle could you rebase to resolve the CI failure? I will take a look afterwards.

mvle · 2024-01-27T02:40:35Z

@Huang-Wei we've rebased. when you get a chance, please provide feedback.

Huang-Wei

Could you eliminate the in-house clientset? We already have mature semantics to interact with any kind of CR objects.

Also, given this PR introduced a CRD, could you also update hack/verify-crdgen.sh to include the source types.go of security-profiles-operator.x-k8s.io_seccompprofiles.yaml?

scheduler-plugins/hack/verify-crdgen.sh

Lines 32 to 33 in c752c54

    
           # Generate CRD 
        
           api_paths="./apis/scheduling/v1alpha1/...;./vendor/github.com/k8stopologyawareschedwg/noderesourcetopology-api/pkg/apis/...;./vendor/github.com/diktyo-io/appgroup-api/pkg/apis/...;./vendor/github.com/diktyo-io/networktopology-api/pkg/apis/..."

Huang-Wei · 2024-01-27T17:55:50Z

...or/sigs.k8s.io/security-profiles-operator/api/seccompprofile/v1beta1/seccompprofile_types.go

could you follow other crds to make it symbolic linked from ../seccompprofile/crd.yaml?:

pr/568 ⇒ ll manifests/crds total 12K lrwxr-xr-x 1 weih 20 Apr 14 2023 appgroup.diktyo.x-k8s.io_appgroups.yaml -> ../appgroup/crd.yaml lrwxr-xr-x 1 weih 27 Apr 14 2023 networktopology.diktyo.x-k8s.io_networktopologies.yaml -> ../networktopology/crd.yaml lrwxr-xr-x 1 weih 30 Apr 14 2023 scheduling.x-k8s.io_elasticquotas.yaml -> ../capacityscheduling/crd.yaml lrwxr-xr-x 1 weih 24 Apr 14 2023 scheduling.x-k8s.io_podgroups.yaml -> ../coscheduling/crd.yaml -rw-r--r-- 1 weih 9.6K Jan 27 09:51 seccompprofiles.security-profiles-operator.x-k8s.io_sysched.yaml lrwxr-xr-x 1 weih 32 May 25 2022 topology.node.k8s.io_noderesourcetopologies.yaml -> ../noderesourcetopology/crd.yaml

Huang-Wei · 2024-01-28T19:12:48Z

test/integration/sysched_test.go

@@ -0,0 +1,322 @@
+/*
+Copyright 2020 The Kubernetes Authors.


Huang-Wei · 2024-01-28T19:25:05Z

pkg/sysched/sysched.go

+// Currently, this function does not change score as the weight is set to 1, and
+// no critical syscalls are given as input
+func (sc *SySched) calcScore(syscalls []string) int {
+	tot_crit := 0


the comment was given on tot_crit := 0, not the function name.

Huang-Wei · 2024-01-28T19:29:55Z

pkg/sysched/clientset/v1alpha1/spo_client.go

+	RestClient rest.Interface
+}
+
+func NewForConfig(c *rest.Config) (*SPOV1Alpha1Client, error) {


We don't need this as this should have been abstracted via generic (dynamic) clientset. You can just use the generic clientset in pkg/sysched/sysched.go w/o composing pkg/sysched/clientset at all.

Example:

scheduler-plugins/pkg/coscheduling/coscheduling.go

Lines 69 to 73 in c752c54

scheme := runtime.NewScheme()

_ = clientscheme.AddToScheme(scheme)

_ = v1.AddToScheme(scheme)

_ = v1alpha1.AddToScheme(scheme)

client, err := client.New(handle.KubeConfig(), client.Options{Scheme: scheme})

Huang-Wei · 2024-01-28T19:31:02Z

test/integration/sysched_test.go

+	err := client.RestClient.
+		Get().
+		Namespace("default").
+		Resource("seccompprofiles").
+		Name(name).
+		VersionedParams(&opts, scheme.ParameterCodec).
+		Do(context.TODO()).
+		Into(&result)


ditto. We don't directly manipulate the client in this manner.

Huang-Wei · 2024-01-28T19:32:56Z

manifests/install/charts/as-a-second-scheduler/templates/rbac.yaml

+- apiGroups: ["security-profiles-operator.x-k8s.io"]
+  resources: ["seccompprofiles", "profilebindings"]
+  verbs: ["get", "list", "watch", "create", "delete", "update", "patch"]


Could you comment it out? for newly introduced plugin, we don't enable it until it's tested enough.

Huang-Wei · 2024-01-28T19:33:46Z

manifests/install/charts/as-a-second-scheduler/templates/rbac.yaml

+- apiGroups: ["security-profiles-operator.x-k8s.io"]
+  resources: ["seccompprofiles", "profilebindings"]
+  verbs: ["get", "list", "watch", "create", "delete", "update", "patch"]


Ditto. Comment them out.

Huang-Wei · 2024-01-28T19:34:14Z

manifests/install/charts/as-a-second-scheduler/values.yaml

@@ -17,7 +17,7 @@ controller:
 # as they need extra RBAC privileges on metrics.k8s.io.

 plugins:
-  enabled: ["Coscheduling","CapacityScheduling","NodeResourceTopologyMatch","NodeResourcesAllocatable"]
+  enabled: ["Coscheduling","CapacityScheduling","NodeResourceTopologyMatch","NodeResourcesAllocatable", "SySched"]


Ditto. Don't enable it by default.

salmanyam · 2024-02-12T18:42:07Z

Hi @Huang-Wei, we have addressed all the comments and removed the in-house clientset. Please take a look and let us know your feedback. Thanks!

salmanyam · 2024-02-27T17:29:34Z

Hi @Huang-Wei, when you get a chance, please provide us feedback.

Huang-Wei

Thanks for the work. It's very close to merge, just a new nits.

BTW: could you squash the commits into 3:

API files (apis/) - manually added by you
API codegen files (apis/../zz_*.go)
Other logic files - i.e., all files exclude 1,2 and 4
vendor/, go.mod, go.sum

Huang-Wei · 2024-03-02T06:16:54Z

manifests/install/charts/as-a-second-scheduler/values.yaml

+- name: SySched
+  args:
+    defaultProfileNamespace: "default"
+    defaultProfileName: "full-seccomp"


Comment these out.

Huang-Wei · 2024-03-02T06:18:25Z

manifests/install/all-in-one.yaml

+- apiGroups: ["security-profiles-operator.x-k8s.io"]
+  resources: ["seccompprofiles", "profilebindings"]
+  verbs: ["get", "list", "watch", "create", "delete", "update", "patch"]


Comment these out (to be consistent with default values.yaml in Helm)

Huang-Wei · 2024-03-02T06:18:34Z

manifests/install/all-in-one.yaml

+- apiGroups: ["security-profiles-operator.x-k8s.io"]
+  resources: ["seccompprofiles", "profilebindings"]
+  verbs: ["get", "list", "watch", "create", "delete", "update", "patch"]


Comment these out (to be consistent with default values.yaml in Helm)

Huang-Wei · 2024-03-02T06:20:02Z

pkg/sysched/sysched.go

+
+type SySched struct {
+	handle framework.Handle
+	//clientSet v1alpha1.SPOV1Alpha1Interface


Let's remove this line.

Huang-Wei · 2024-03-02T06:20:12Z

pkg/sysched/sysched.go

+	// NOTE: not used at this time
+	// Key: category of critical system call, e.g., CVE based, admin policy, etc.
+	// Value: list of system calls
+	//CritSyscalls            map[string][]string


Ditto. Remove this line.

Huang-Wei · 2024-03-02T06:20:28Z

pkg/sysched/sysched.go

+		// SCMP_ACT_TRACE --> ActTrace, seccomp.ActNotify
+		if element.Action == seccomp.ActAllow || element.Action == seccomp.ActLog {
+			syscalls = syscalls.Union(sets.New[string](element.Names...))
+			//syscalls = unionList(syscalls, element.Names)


Ditto. Let's remove this line.

Huang-Wei · 2024-03-02T06:21:47Z

pkg/sysched/sysched.go

+	// NOTE: weight W is hardcoded for now
+	// TODO: add critical/cve syscalls
+	// TODO: adjust weight W for critical/cve syscalls
+	tot_crit := 0


Could you change it to totCrit? to be Go-idiomatic.

Huang-Wei · 2024-03-02T06:22:42Z

pkg/sysched/sysched.go

+		return 0, nil
+	}
+
+	//diffSyscalls := setSubtract(hostSyscalls, podSyscalls)


Remove this line please.

Manually update the API files for the implementation of SySched, a system call-aware scheduler(KEP kubernetes-sigs#399). SySched is a new scoring plugin that ranks feasible nodes based on the relative risks of pods' system call usage.

salmanyam · 2024-03-06T15:08:20Z

Hi @Huang-Wei, we made the changes accordingly. Please take a look when you a chance. Thanks a lot!

Huang-Wei · 2024-03-06T17:34:16Z

@salmanyam thanks!

/lgtm
/approve

k8s-ci-robot · 2024-03-06T17:34:30Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: Huang-Wei, salmanyam

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

~~OWNERS~~ [Huang-Wei]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

salmanyam · 2024-03-06T17:38:47Z

Thanks a lot, @Huang-Wei! We greatly appreciate your time and effort on this PR.

k8s-ci-robot added release-note Denotes a PR that will be considered when it comes time to generate release notes. kind/feature Categorizes issue or PR as related to a new feature. labels Mar 30, 2023

k8s-ci-robot added cncf-cla: no Indicates the PR's author has not signed the CNCF CLA. needs-ok-to-test Indicates a PR that requires an org member to verify it is safe to test. do-not-merge/contains-merge-commits labels Mar 30, 2023

k8s-ci-robot requested review from PiotrProkop and swatisehgal March 30, 2023 16:18

k8s-ci-robot added the size/XXL Denotes a PR that changes 1000+ lines, ignoring generated files. label Mar 30, 2023

salmanyam force-pushed the sysched-scoring branch from 5ec6eb3 to ca6eee7 Compare March 30, 2023 17:43

k8s-ci-robot added cncf-cla: yes Indicates the PR's author has signed the CNCF CLA. and removed do-not-merge/contains-merge-commits cncf-cla: no Indicates the PR's author has not signed the CNCF CLA. labels Mar 30, 2023

salmanyam changed the title ~~Sysched: add the implementation of SySched (KEP #399)~~ Sysched: add the implementation of SySched Mar 31, 2023

PiotrProkop suggested changes Apr 11, 2023

View reviewed changes

k8s-ci-robot added the do-not-merge/contains-merge-commits label Apr 19, 2023

k8s-ci-robot added ok-to-test Indicates a non-member PR verified by an org member that is safe to test. and removed needs-ok-to-test Indicates a PR that requires an org member to verify it is safe to test. labels Apr 21, 2023

k8s-ci-robot added the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label May 19, 2023

salmanyam force-pushed the sysched-scoring branch from 214e36c to 5265676 Compare May 19, 2023 20:13

k8s-ci-robot removed needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. do-not-merge/contains-merge-commits labels May 19, 2023

k8s-ci-robot added the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label Jun 2, 2023

salmanyam force-pushed the sysched-scoring branch from 49100dc to 2da5026 Compare June 15, 2023 20:07

Huang-Wei reviewed Sep 18, 2023

View reviewed changes

salmanyam force-pushed the sysched-scoring branch from 4e9cd1e to 871e7d1 Compare October 24, 2023 15:39

k8s-ci-robot removed the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label Oct 28, 2023

k8s-ci-robot added the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label Nov 10, 2023

salmanyam force-pushed the sysched-scoring branch from 22dad6e to bf69b05 Compare January 18, 2024 19:09

k8s-ci-robot removed the needs-rebase Indicates a PR cannot be merged because it has merge conflicts with HEAD. label Jan 27, 2024

Huang-Wei requested changes Jan 28, 2024

View reviewed changes

Huang-Wei reviewed Mar 2, 2024

View reviewed changes

salmanyam force-pushed the sysched-scoring branch from 8ae9391 to f3c4d5a Compare March 5, 2024 21:21

salmanyam added 4 commits March 5, 2024 15:27

Update API files

41473cb

Manually update the API files for the implementation of SySched, a system call-aware scheduler(KEP kubernetes-sigs#399). SySched is a new scoring plugin that ranks feasible nodes based on the relative risks of pods' system call usage.

Update API codegen files

1295000

Add other logic files

588b8ec

Update vendor, go.mod, and go.sum

4b6ea59

salmanyam force-pushed the sysched-scoring branch from f3c4d5a to 4b6ea59 Compare March 5, 2024 21:28

k8s-ci-robot assigned Huang-Wei Mar 6, 2024

k8s-ci-robot added the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Mar 6, 2024

k8s-ci-robot added the approved Indicates a PR has been approved by an approver from all required OWNERS files. label Mar 6, 2024

k8s-ci-robot merged commit cd3e4fb into kubernetes-sigs:master Mar 6, 2024
10 checks passed

		#serviceAccountName: default
		#schedulerName: scheduler-plugins-scheduler


		nodeIP := node.Status.Addresses[0].Address

		podSyscalls, _ := sc.getSyscalls(pod)

	# Generate CRD
	api_paths="./apis/scheduling/v1alpha1/...;./vendor/github.com/k8stopologyawareschedwg/noderesourcetopology-api/pkg/apis/...;./vendor/github.com/diktyo-io/appgroup-api/pkg/apis/...;./vendor/github.com/diktyo-io/networktopology-api/pkg/apis/..."

	scheme := runtime.NewScheme()
	_ = clientscheme.AddToScheme(scheme)
	_ = v1.AddToScheme(scheme)
	_ = v1alpha1.AddToScheme(scheme)
	client, err := client.New(handle.KubeConfig(), client.Options{Scheme: scheme})

Sysched: add the implementation of SySched #568

Sysched: add the implementation of SySched #568

Conversation

salmanyam commented Mar 30, 2023

What type of PR is this?

What this PR does / why we need it:

Which issue(s) this PR fixes:

Special notes for your reviewer:

Does this PR introduce a user-facing change?

linux-foundation-easycla bot commented Mar 30, 2023 • edited Loading

k8s-ci-robot commented Mar 30, 2023

salmanyam commented Mar 30, 2023

Huang-Wei commented Apr 5, 2023

PiotrProkop left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Huang-Wei commented Apr 21, 2023

salmanyam commented Apr 21, 2023

salmanyam commented May 22, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mvle commented Sep 21, 2023

netlify bot commented Oct 24, 2023 • edited Loading

✅ Deploy Preview for kubernetes-sigs-scheduler-plugins canceled.

mvle commented Nov 2, 2023

mvle commented Jan 17, 2024

Huang-Wei commented Jan 18, 2024

mvle commented Jan 27, 2024

Huang-Wei left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

salmanyam commented Feb 12, 2024

salmanyam commented Feb 27, 2024

Huang-Wei left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

salmanyam commented Mar 6, 2024

Huang-Wei commented Mar 6, 2024

k8s-ci-robot commented Mar 6, 2024

salmanyam commented Mar 6, 2024

linux-foundation-easycla bot commented Mar 30, 2023 •

edited

Loading

netlify bot commented Oct 24, 2023 •

edited

Loading