mirrors/node-feature-discovery
mirrors/node-feature-discovery
1
0
Fork 0
mirror of https://github.com/kubernetes-sigs/node-feature-discovery.git synced 2025-04-15 00:36:23 +00:00
Code Activity
2909 commits 14 branches 71 tags 139 MiB
Commit graph

481 commits

Author SHA1 Message Date
Kubernetes Prow Robot
c6c517ed8a
Merge pull request #2055 from routerhan/feat-add-hypervisor 
feat: add support to differentiate specific hypervisors on s390x
 2025-04-10 02:30:53 -07:00
Kubernetes Prow Robot
d9db00d345
Merge pull request #2096 from mfranczy/version-range 
Allow to define types for NFR/NFG expressions
 2025-04-07 00:20:39 -07:00
Ching Han Chen
8161e6e62e feat: add support to differentiate specific hypervisors on s390x 
Signed-off-by: Ching Han Chen <ching.chen@ibm.com>
 2025-04-04 09:15:30 +02:00
Marcin Franczyk
4cf8032ac5 NFD API: Add the 'type' field for MatchExpression 2025-04-03 20:16:03 +02:00
Markus Lehtonen
225070eeab nfd-master: drop the deprecated autoDefaultNs config option 
The autoDefaultNs was deprecated in NFD v0.16 and replaced by the
DisableAutoPrefix feature gate.
 2025-04-03 10:31:12 +03:00
Oleg Zhurakivskyy
8df8d6c77a Document AMXCOMPLEX, AMXTF32 cpuid features 
Signed-off-by: Oleg Zhurakivskyy <oleg.zhurakivskyy@intel.com>
 2025-03-31 14:12:18 +03:00
Kubernetes Prow Robot
3b5276dc66
Merge pull request #2044 from mfranczy/network-mtu 
Discover MTU of physical and virtual network devices
 2025-03-27 09:14:39 -07:00
Markus Lehtonen
7a8d3efce3 docs: bump docs build deps 
Fix docs build failure on latest ruby:3.1 docker image.
 2025-03-27 14:37:26 +02:00
Kubernetes Prow Robot
6f6f5d68e1
Merge pull request #2087 from marquiz/devel/compat-api-types 
compat/api: use GroupRule type
 2025-03-14 12:35:51 -07:00
Markus Lehtonen
fba7fb6c94 compat/api: use GroupRule type 
Simplify and rationalize the compat api. Many of the fields in Rule,
like Labels, Annotations or ExtendedResources are not applicable in this
context.
 2025-03-14 13:40:57 +02:00
Marcin Franczyk
7ff00b3a02 Add Ge Le and GeLe operators 
Signed-off-by: Marcin Franczyk <marcin0franczyk@gmail.com>
 2025-03-04 12:56:53 +01:00
Kubernetes Prow Robot
e7deb5cc96
Merge pull request #2056 from mfranczy/hugepages 
Introduce memory.hugepages feature
 2025-03-03 02:33:17 -08:00
Marcin Franczyk
bbff335fb7 Update huge pages documentation 
Signed-off-by: Marcin Franczyk <marcin0franczyk@gmail.com>
 2025-03-03 10:14:05 +01:00
Marcin Franczyk
abbcf94dd6 Introduce memory.hugepages feature 
Signed-off-by: Marcin Franczyk <marcin0franczyk@gmail.com>
 2025-02-24 09:43:07 +01:00
Fahrzin Hemmati
b0dbd11af7 Use Sprig template functions and add asLabelValue function 2025-02-19 01:51:32 -08:00
Kubernetes Prow Robot
797c66ffaa
Merge pull request #1947 from marquiz/devel/health-topology-updater 
nfd-topology-updater: serve metrics and healthz on the same port
 2025-02-17 02:02:22 -08:00
Marcin Franczyk
69f5009100 Add MTU attribute for physical and virtual devices 
Signed-off-by: Marcin Franczyk <marcin0franczyk@gmail.com>
 2025-02-13 15:57:35 +01:00
Markus Lehtonen
0b0aed2318 nfd-master: serve metrics and healthz endpoint on the same port 
Changes the gRPC health endpoint to plain http. At the same time starts
serving both the metrics and healthz endpoints on a single port.
Replaces the -metrics and -grpc-health command line flags with a single
-port flag.

Changes the Helm and kustomize deployments correspondingly.
 2025-02-04 10:30:58 +02:00
Markus Lehtonen
0b9a8cf120 nfd-topology-updater: serve metrics and healthz on the same port 
Changes the gRPC health endpoint to plain http. At the same time starts
serving both the metrics and healthz endpoints on a single port.
Replaces the -metrics and -grpc-health command line flags with a single
-port flag.

Changes the Helm and kustomize deployments correspondingly.
 2025-02-04 10:30:26 +02:00
Markus Lehtonen
4959a13a07 nfd-worker: replace --metrics with --port 
Use a single port for serving http. In addition to metrics we will have
the healthz endpoint.
 2025-01-31 07:52:12 +02:00
Markus Lehtonen
25914ec06e nfd-worker: drop the gRPC health port 
To be replaced with plain http.
 2025-01-31 07:50:00 +02:00
killianmuldoon
0ff1987d2a
Make DNS Policy configurable in helm chart (#2025) 
* Make DNS Policy configurable in helm chart

Signed-off-by: Killian Muldoon <kmuldoon@nvidia.com>

* Add documentation for helm dnsPolicy values

Signed-off-by: Killian Muldoon <kmuldoon@nvidia.com>

* Reformat tables in helm docs

Signed-off-by: Killian Muldoon <kmuldoon@nvidia.com>

---------

Signed-off-by: Killian Muldoon <kmuldoon@nvidia.com>
 2025-01-23 01:42:58 -08:00
Marcin Franczyk
2208978390
Update the documentation 
Signed-off-by: Marcin Franczyk <marcin0franczyk@gmail.com>
 2024-12-18 20:23:52 +01:00
Marcin Franczyk
0b7661bf17
Add experimental note and fix subcmds flags naming 
Signed-off-by: Marcin Franczyk <marcin0franczyk@gmail.com>
 2024-12-18 15:39:18 +01:00
Marcin Franczyk
5b57312d74 Add image compatibility documentation 
Signed-off-by: Marcin Franczyk <marcin0franczyk@gmail.com>
 2024-12-18 10:49:02 +01:00
Kubernetes Prow Robot
3e87c97ac2
Merge pull request #1976 from marquiz/devel/grpc-api-cleanup 
Cleanup for NodeFeature API being GA
 2024-12-13 15:14:26 +01:00
Markus Lehtonen
fc103a6028 Cleanup for NodeFeature API being GA 
Drop references to the gRPC API and don't suggest that NodeFeatureAPI
could be disabled.

Also update the developer guide for instructions running nfd components
outside the cluster.
 2024-12-13 15:40:46 +02:00
Kubernetes Prow Robot
caaac59eba
Merge pull request #1860 from ozhuraki/no-owner-refs 
nfd-worker: Add an option to disable setting the owner references
 2024-12-13 13:12:26 +01:00
Markus Lehtonen
fb6484fb8d deployment: add startupProbe for nfd-master 
This patch mitigates inadvertent termination of nfd-master pods by the
liveness probe on big clusters.

With a recent change nfd-master started to wait (block) for informer
caches to sync before starting the main loop. Consequently, this change
also made the gRPC health enpoint to not respond until the caches have
been synced. In big clusters the syncing the NodeFeature object cache
takes a long time as the objects are big and there's (at least) one per
each node in the cluster. Thus, in big clusters, the liveness probe
kicks in and kills the nfd-master pod before it's ready.
 2024-12-12 20:00:49 +02:00
Oleg Zhurakivskyy
20ef877ab1 nfd-worker: Add an option to disable setting the owner references 
In some cases it's desirable to control automatic garbage collection
of NodeFeature object.

Add an option to disable setting the owner references to Pod
for NodeFeature object.

Closes: 1817

Signed-off-by: Oleg Zhurakivskyy <oleg.zhurakivskyy@intel.com>
 2024-11-28 16:50:10 +02:00
Kubernetes Prow Robot
443913e019
Merge pull request #1956 from googs1025/chore/add_metrics_prefix 
chore: add metrics system prefix
 2024-11-28 09:00:59 +00:00
googs1025
e631a52374 chore: add metrics system prefix 2024-11-28 09:57:40 +08:00
Oleg Zhurakivskyy
fb52206b96 Detect AMXFP8 cpuid feature 
Signed-off-by: Oleg Zhurakivskyy <oleg.zhurakivskyy@intel.com>
 2024-11-26 14:39:18 +02:00
Kubernetes Prow Robot
835832729f
Merge pull request #1951 from marquiz/devel/nodefeature-ga 
docs: minor update in the feature gates table
 2024-11-08 08:50:43 +00:00
Markus Lehtonen
1244a42030 docs: minor update in the feature gates table 2024-11-07 15:27:51 +02:00
Markus Lehtonen
45f49d574a nfd-master: drop resourceLabels 
Drop the resourceLabels config file option and the corresponding
-resource-labels command line flag. They were deprecated in NFD v0.13 so
it's time to let them go. NodeFeatureRule(s) should be used to manage
ERs, instead.
 2024-11-07 15:16:52 +02:00
Kubernetes Prow Robot
61ce3b3ce3
Merge pull request #1948 from marquiz/devel/deprecate-separate-ports 
Deprecate separate metrics and health port args
 2024-11-06 11:05:29 +00:00
Markus Lehtonen
4bb91e2096 Deprecate separate metrics and health port args 2024-11-06 12:14:41 +02:00
Kubernetes Prow Robot
955095c7eb
Merge pull request #1889 from ChaoyiHuang/fixtiltup 
Doc: Fix tilt up issue in feature discovering in developer guide
 2024-11-06 10:03:30 +00:00
Carlos Eduardo Arango Gutierrez
62f4eddce6
Drop support for hooks 
Signed-off-by: Carlos Eduardo Arango Gutierrez <eduardoa@nvidia.com>
 2024-11-04 14:50:07 +01:00
Kubernetes Prow Robot
65b5e0c255
Merge pull request #1944 from ArangoGutierrez/I/1733 
Taints: mark stable
 2024-11-04 12:49:29 +00:00
Carlos Eduardo Arango Gutierrez
dc7edd50ba
Taints: mark stable 
Signed-off-by: Carlos Eduardo Arango Gutierrez <eduardoa@nvidia.com>
 2024-11-04 11:30:09 +01:00
Kubernetes Prow Robot
b997ade5b3
Merge pull request #1942 from marquiz/devel/drop-grpc 
nfd-master: drop stale unreachable deprecation notices
 2024-11-04 11:16:31 +01:00
Chaoyi Huang
d08ea5ee11 Doc: Fix tilt up issue in feature discovering in developer guide 
The issue is due to the k3d/kind cluster created by ctlptl will run
inside containers(it will serve as the virtual hosts).

Host folders which will be scaned by the nfd feature discovery should
be mounted into the container ( the virtual host). otherwise the nfd-worker
container which run inside the virtual host will just see the default base
image rootfs /boot, /lib folders, which are usually empty, leads to the
discovey failure.

Signed-off-by: Chaoyi Huang <joehuang.sweden@gmail.com>
 2024-11-01 02:31:23 +00:00
Kubernetes Prow Robot
fd2893e2a5
Merge pull request #1592 from AhmedThresh/feat-configure-cr-restrictions 
feat/nfd-master: configure CR restrictions
 2024-10-24 12:20:54 +01:00
Tobias Giese
52c2fc6498
Add separate helm values for the liveness and readiness probes 
Signed-off-by: Tobias Giese <tgiese@nvidia.com>
 2024-10-18 12:54:42 +02:00
Tobias Giese
901fbe2866
Format helm.md 
Signed-off-by: Tobias Giese <tgiese@nvidia.com>
 2024-10-18 12:54:42 +02:00
Markus Lehtonen
010393b302 docs: quote shell snippets containing urls with query parameters 
Makes them work with zsh which tries to glob URLs containing query
parameters (question marks).
 2024-10-02 17:07:32 +03:00
Tobias Giese
53ddf081da
Add parameter to configure health endpoint port 
Signed-off-by: Tobias Giese <tgiese@nvidia.com>
 2024-09-24 15:15:50 +02:00
Tobias Giese
af0592b87c
Add helm values to configure hostNetwork and additional env vars 
We have to run our NFD workers in the host network.
Also we need additional env variables such as KUBERNETES_SERVICE_HOST and _PORT.
To achieve this we can simply add generic helm values. The default behavior is not changed.

Signed-off-by: Tobias Giese <tgiese@nvidia.com>
 2024-09-18 17:58:59 +02:00