manager: sync rdma resource to node #2249
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ZiMengSheng
force-pushed
the
koord-manager-rdma
branch
2 times, most recently
from
2fa36c7 to
b1c9e60
Compare
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #2249 +/- ##
==========================================
+ Coverage 66.07% 66.10% +0.02%
==========================================
Files 454 456 +2
Lines 53451 53632 +181
==========================================
+ Hits 35318 35451 +133
- Misses 15593 15632 +39
- Partials 2540 2549 +9
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
ZiMengSheng
force-pushed
the
koord-manager-rdma
branch
from
b1c9e60 to
8cafd84
Compare
|
/lgtm |
ZiMengSheng
changed the title
saintube
reviewed
Dec 4, 2024
pkg/slo-controller/noderesource/plugins/gpudeviceresource/plugin.go
Outdated
Show resolved
Hide resolved
pkg/slo-controller/noderesource/plugins/rdmadevicereource/plugin.go
Outdated
Show resolved
Hide resolved
pkg/slo-controller/noderesource/plugins/rdmadevicereource/plugin.go
Outdated
Show resolved
Hide resolved
ZiMengSheng
force-pushed
the
koord-manager-rdma
branch
from
8cafd84 to
528e095
Compare
saintube
reviewed
Dec 5, 2024
pkg/slo-controller/noderesource/plugins/gpudeviceresource/plugin.go
Outdated
Show resolved
Hide resolved
pkg/slo-controller/noderesource/plugins/rdmadevicereource/plugin.go
Outdated
Show resolved
Hide resolved
Signed-off-by: wangjianyu.wjy <[email protected]>
ZiMengSheng
force-pushed
the
koord-manager-rdma
branch
from
528e095 to
30a4b7a
Compare
|
/approve |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
[koord-manager-rdma]
Ⅰ. Describe what this PR does
In order to implement the end-to-end scheme of the rdma device, it is necessary to add an rdma resource controller to update the status of node nodes in time, including Capacity and Allocatable. The koordinator.sh/rdma resources are mainly updated.
Ⅱ. Does this pull request fix one issue?
Fixed loss of functionality for rdma resource registration and update to nodes
Ⅲ. Describe how to verify it
In the k8s cluster, prepare one or more servers that support rdma network adapters as cluster nodes. Install the new version of the koordlet component on each node and install the new version of the koord-manager component. Check whether the number of rdma resources in a node is updated to the actual number of rdma nics on the node.
Uninstall one or more RDMA network cards of the node and restart the server. After the node is added to the cluster and the cluster is stable, check whether the number of rdma network cards in the node resource is updated to the actual number of rdma network cards of the node.
Ⅳ. Special notes for reviews
Updating RDMA device resources to nodes requires the koordlet component on the end. Therefore, you must install the koordlet version that supports RDMA on each node in the cluster. In this way, the koord-manager component can be used to test the effect
V. Checklist
make test