Skip to content
This repository has been archived by the owner on Mar 6, 2024. It is now read-only.

Commit

Permalink
0.4-update (#53)
Browse files Browse the repository at this point in the history
* Update README.md

* Create README.md

* Update README.md

* Update README.md

* fix-secret-namespace

* Update README.md

* Update README.md

* Update README.md

* Add files via upload

* Update README.md

* Update util.go

* Create build.yml

* Update build.yml

* Delete .github directory

* Update Dockerfile

* Create docker-image.yml

* Update docker-image.yml

* Update docker-image.yml

* Update docker-image.yml

* Update README.md

* Update README.md

* Delete docker-image.yml

* Update README.md

* Update README.md

* Update Makefile

* Update bitfusion-client-configmap.yaml

* Update pod.yaml

* Update deployment.yaml

* Update job.yaml

* Update pod-filter.yaml

* Update pod-memory.yaml

Co-authored-by: root <[email protected]>
  • Loading branch information
ln23415 and root authored Dec 9, 2021
1 parent 0fa882d commit 4fcca3b
Show file tree
Hide file tree
Showing 8 changed files with 46 additions and 46 deletions.
2 changes: 1 addition & 1 deletion bitfusion_device_plugin/Makefile
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@ IMAGE_REPO ?= docker.io/bitfusiondeviceplugin
DEVICE_IMAGE_NAME ?= bitfusion-device-plugin
WEBHOOK_IMAGE_NAME ?= bitfusion-webhook
PKG_IMAGE_NAME ?= bitfusion-client
IMAGE_TAG ?= 0.3
IMAGE_TAG ?= 0.4
K8S_PLATFORM ?= community

export K8S_PLATFORM
Expand Down
40 changes: 20 additions & 20 deletions bitfusion_device_plugin/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -64,8 +64,8 @@ bitfusion-webhook runs as a Deployment on the Kubernetes master node.
## 2. Prerequisites
- Ubuntu Linux as the operating system of the installation machine
- OpenSSL needs to be installed on Ubuntu
- Kubernetes 1.17+
- Bitfusion 3.5 and 4.0.1
- Kubernetes 1.16+
- Bitfusion 4.0.1 and 4.5.0
- kubectl and docker command are ready to use.
- need to specify the command field in the POD

Expand Down Expand Up @@ -201,7 +201,7 @@ IMAGE_REPO ?= docker.io/bitfusiondeviceplugin
DEVICE_IMAGE_NAME ?= bitfusion-device-plugin
WEBHOOK_IMAGE_NAME ?= bitfusion-webhook
PKG_IMAGE_NAME ?= bitfusion-client
IMAGE_TAG ?= 0.3
IMAGE_TAG ?= 0.4
```

Now start building images using the command below:
Expand All @@ -214,9 +214,9 @@ If everything works well, use the following command to check images:
```shell
$ docker images
REPOSITORY TAG
docker.io/bitfusiondeviceplugin/bitfusion-device-plugin 0.3
docker.io/bitfusiondeviceplugin/bitfusion-webhook 0.3
docker.io/bitfusiondeviceplugin/bitfusion-client 0.3
docker.io/bitfusiondeviceplugin/bitfusion-device-plugin 0.4
docker.io/bitfusiondeviceplugin/bitfusion-webhook 0.4
docker.io/bitfusiondeviceplugin/bitfusion-client 0.4

```

Expand Down Expand Up @@ -310,7 +310,7 @@ After completing the installation, users can write a YAML file of Kubernetes to
| bitfusion.io/gpu-percent | positive integer |Percentage of the memory of each GPU|
| bitfusion.io/gpu-memory | positive integer |Memory size of each GPU,The default unit is bit.It can be used with the K8s native memory application unit (Mi,M,G,Gi)|
| bitfusion-client/os | ubuntu18 / ubuntu20 / centos7 / centos8 |The OS of the containers that use the Bitfusion client|
| bitfusion-client/version | 350/401 |The version of Bitfusion client to be used in this container is 3.5 or 4.0.1|
| bitfusion-client/version | 401/450 |The version of Bitfusion client to be used in this container is 3.5 or 4.0.1|


Below is a sample YAML of Pod which runs a benchmark of Tensorflow. The variable `hostPath` is the directory where the Tensorflow Benchmarks code resides on the host and it will be mounted into the pod.
Expand All @@ -328,7 +328,7 @@ metadata:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
name: bf-pkgs
# You can specify any namespace
namespace: tensorflow-benchmark
Expand Down Expand Up @@ -376,7 +376,7 @@ metadata:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
name: bf-pkgs
# You can specify any namespace
namespace: tensorflow-benchmark
Expand Down Expand Up @@ -500,7 +500,7 @@ total images/sec: 199.65

### 4.3. The configuration of "auto-management/bitfusion parameter"

If the value of auto-management/bitfusion parameter is set to "all", use the following command to to check the status of the pod when the workload has been summited.
If the value of auto-management/bitfusion parameter is set to "all", use the following command to to check the status of the pod when the workload has been submitted.

```bash
$ kubectl edit pod -n tensorflow-benchmark bf-pkgs
Expand Down Expand Up @@ -610,7 +610,7 @@ metadata:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
# Use a single filter condition
bitfusion-client/filter: "server.hostname=bf-server"
name: bf-pkgs
Expand Down Expand Up @@ -643,7 +643,7 @@ metadata:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
# Use multiple filter criteria
bitfusion-client/filter: "server.hostname=bf-server server.addr=192.168.1.1 server.hostname=bf-server2"
name: bf-pkgs
Expand Down Expand Up @@ -708,7 +708,7 @@ metadata:
  annotations:
    auto-management/bitfusion: "all"
    bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
  name: bf-pkgs
  namespace: tensorflow-benchmark
spec:
Expand Down Expand Up @@ -762,7 +762,7 @@ metadata:
  annotations:
    auto-management/bitfusion: "all"
    bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
  name: bf-pkgs
  namespace: tensorflow-benchmark
spec:
Expand Down Expand Up @@ -917,18 +917,18 @@ After that run the "make deploy" command to start the deployment.
Below is another image regsitry for users to get the docker images.
- Option 1: Tencentyun
```shell
ccr.ccs.tencentyun.com/bitfusion/bitfusion-device-plugin:0.3
ccr.ccs.tencentyun.com/bitfusion/bitfusion-device-plugin:0.4

ccr.ccs.tencentyun.com/bitfusion/bitfusion-webhook:0.3
ccr.ccs.tencentyun.com/bitfusion/bitfusion-webhook:0.4

ccr.ccs.tencentyun.com/bitfusion/bitfusion-client:0.3
ccr.ccs.tencentyun.com/bitfusion/bitfusion-client:0.4
```
- Option 2: GHCR
```
ghcr.io/ln23415/bitfusion-device-plugin:0.3
ghcr.io/ln23415/bitfusion-device-plugin:0.4
ghcr.io/ln23415/bitfusion-webhook:0.3
ghcr.io/ln23415/bitfusion-webhook:0.4
ghcr.io/ln23415/bitfusion-client:0.3
ghcr.io/ln23415/bitfusion-client:0.4
```

2 changes: 1 addition & 1 deletion bitfusion_device_plugin/example/deployment.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@ spec:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
name: bf-pkgs
labels:
app: share-gpu
Expand Down
2 changes: 1 addition & 1 deletion bitfusion_device_plugin/example/job.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -9,7 +9,7 @@ spec:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
name: bf-pkgs
spec:
restartPolicy: OnFailure
Expand Down
2 changes: 1 addition & 1 deletion bitfusion_device_plugin/example/pod-filter.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ metadata:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
bitfusion-client/filter: "server.hostname=bf-server"
name: bf-pkgs
namespace: tensorflow-benchmark
Expand Down
2 changes: 1 addition & 1 deletion bitfusion_device_plugin/example/pod-memory.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ metadata:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
name: bf-pkgs
namespace: tensorflow-benchmark
spec:
Expand Down
2 changes: 1 addition & 1 deletion bitfusion_device_plugin/example/pod.yaml
Original file line number Diff line number Diff line change
Expand Up @@ -4,7 +4,7 @@ metadata:
annotations:
auto-management/bitfusion: "all"
bitfusion-client/os: "ubuntu18"
bitfusion-client/version: "401"
bitfusion-client/version: "450"
name: bf-pkgs
namespace: tensorflow-benchmark
spec:
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -7,6 +7,26 @@ data:
bitfusion-client-config.yaml: |
BitfusionClients:
- BitfusionVersion: "450"
OSVersion: ubuntu18
BinaryPath: /bitfusion/bitfusion-client-ubuntu1804_4.5.0-4_amd64.deb/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-ubuntu1804_4.5.0-4_amd64.deb/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "450"
OSVersion: ubuntu20
BinaryPath: /bitfusion/bitfusion-client-ubuntu2004_4.5.0-4_amd64.deb/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-ubuntu2004_4.5.0-4_amd64.deb/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "450"
OSVersion: centos7
BinaryPath: /bitfusion/bitfusion-client-centos7-4.5.0-4.x86_64.rpm/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-centos7-4.5.0-4.x86_64.rpm/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "450"
OSVersion: centos8
BinaryPath: /bitfusion/bitfusion-client-centos8-4.5.0-4.x86_64.rpm/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-centos8-4.5.0-4.x86_64.rpm/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "401"
OSVersion: ubuntu18
BinaryPath: /bitfusion/bitfusion-client-ubuntu1804_4.0.1-5_amd64.deb/usr/bin/bitfusion
Expand All @@ -26,23 +46,3 @@ data:
OSVersion: centos8
BinaryPath: /bitfusion/bitfusion-client-centos8-4.0.1-5.x86_64.rpm/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-centos8-4.0.1-5.x86_64.rpm/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "350"
OSVersion: ubuntu18
BinaryPath: /bitfusion/bitfusion-client-ubuntu1804_3.5.0-5_amd64.deb/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-ubuntu1804_3.5.0-5_amd64.deb/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "350"
OSVersion: ubuntu20
BinaryPath: /bitfusion/bitfusion-client-ubuntu2004_3.5.0-5_amd64.deb/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-ubuntu2004_3.5.0-5_amd64.deb/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "350"
OSVersion: centos7
BinaryPath: /bitfusion/bitfusion-client-centos7-3.5.0-5.x86_64.rpm/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-centos7-3.5.0-5.x86_64.rpm/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/
- BitfusionVersion: "350"
OSVersion: centos8
BinaryPath: /bitfusion/bitfusion-client-centos8-3.5.0-5.x86_64.rpm/usr/bin/bitfusion
EnvVariable: /bitfusion/bitfusion-client-centos8-3.5.0-5.x86_64.rpm/opt/bitfusion/lib/x86_64-linux-gnu/bitfusion/lib/

0 comments on commit 4fcca3b

Please sign in to comment.