GPU/TPU-resource-usage-collect-addon #20
Conversation
z1ens
changed the title
|
/assign @zhujian7 @haoqing0110 @qiujian16 |
z1ens
force-pushed
the
gpu-tpu-addon-template
branch
2 times, most recently
from
caa40ee
to
83f79ab
Compare
| } | ||
|
|
||
| func (s *Score) calculateClusterAllocateable(resourceName clusterv1.ResourceName) (float64, error) { | ||
| // Iterate every node, find the node with maximum allocatable resource, return the number and node name. | ||
| func (s *Score) calculateClusterAllocatable(resourceName string) (float64, string, error) { |
There was a problem hiding this comment.
calculateMaxAllocatableNode?
|
|
||
| After the deployment is complete, addon will create an addonplacementscore in its own namespace for each managedcluster in the hub. | ||
| On the hub cluster, you can see the `addonTemplate`, and check the `managedClusterAddon` status. |
There was a problem hiding this comment.
AddOnTemplate, ManagedClusterAddon.
There was a problem hiding this comment.
Yes I have changed that.
| }, | ||
| { | ||
| Name: "tpuAvailable", | ||
| Value: tpuScore, | ||
| }, |
There was a problem hiding this comment.
Hmm, I'm thinking should we keep the current xxxAvailable as cluster level score and add a new score based on max available node.
There was a problem hiding this comment.
Or at least explain the score is based on the cluster max available node in the readme.
There was a problem hiding this comment.
Yes I should add an explanation in the Documentation.
For an extra score, I think we can do that too, like for example for each cluster, we have cpuClusterAvailable, cpuNodeAvaliable, gpuClusterAvaliable, gpuNodeAvaliable, etc.
| } | ||
|
|
||
| func (s *Score) calculateNodeResourceUsage(nodeName string, resourceName string) (float64, error) { | ||
| list, err := s.podLister.List(labels.Everything()) |
There was a problem hiding this comment.
Completed pods should be filtered out. Eg,
NAME READY STATUS RESTARTS AGE
demo1-job8n84t-bgzhs 0/1 Completed 0 19m
|
@qiujian16 could you help take a final review? |
|
/approve |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: qiujian16, z1ens 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:
Approvers can indicate their approval by writing |
…ach managed clusters, develop the add-on in addontemplate mode. Signed-off-by: z1ens <xxtale02591@gmail.com>
z1ens
force-pushed
the
gpu-tpu-addon-template
branch
from
d748675
to
6503ea3
Compare
|
/lgtm |
|
/unhold |
1 similar comment
|
/unhold |
z1ens
changed the title
openshift-merge-bot
bot
merged commit ebebf50
into
open-cluster-management-io:main
…ach managed clusters, develop the add-on in addontemplate mode. (open-cluster-management-io#20) Signed-off-by: z1ens <xxtale02591@gmail.com>
…ach managed clusters, develop the add-on in addontemplate mode. (open-cluster-management-io#20) Signed-off-by: z1ens <xxtale02591@gmail.com>
templatetag.cluster-roleandaddon-template.placementin a separated file.roleandrolebindinginaddontemplate.goUpdate on Aug.15th:
REF: open-cluster-management-io/ocm#369
Old PR: #16 (comment)