Skip to content

Commit c916e5f

Browse files
author
Anurag Guda
authored
Merge pull request #127 from NVIDIA/25.7.1
25.7.1 Release
2 parents ef0c519 + e995f38 commit c916e5f

20 files changed

+123
-3283
lines changed

‎README.md‎

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -26,17 +26,17 @@ When NVIDIA Cloud Native Stack batch is released, the previous batch enters main
2626

2727
| Batch | Status |
2828
| :-----: | :--------------:|
29-
| [25.7.0](https://github.com/NVIDIA/cloud-native-stack/releases/tag/v25.7.0) | Generally Available |
30-
| [25.4.0](https://github.com/NVIDIA/cloud-native-stack/releases/tag/v25.4.0) | Maintenance |
31-
| [24.11.2](https://github.com/NVIDIA/cloud-native-stack/releases/tag/v24.11.2) | EOL |
29+
| [25.7.1](https://github.com/NVIDIA/cloud-native-stack/releases/tag/v25.7.1) | Generally Available |
30+
| [25.7.0](https://github.com/NVIDIA/cloud-native-stack/releases/tag/v25.7.0) | Maintenance |
31+
| [25.4.0](https://github.com/NVIDIA/cloud-native-stack/releases/tag/v25.4.0) | EOL |
3232

3333
`NOTE:` CNS Version 15.0 and above is Now supports Ubuntu 24.04
3434

3535
For more information, Refer [Cloud Native Stack Releases](https://github.com/NVIDIA/cloud-native-stack/releases)
3636

3737
## Component Matrix
3838

39-
#### Cloud Native Stack Batch 25.7.0 (Release Date: 21 July 2025)
39+
#### Cloud Native Stack Batch 25.7.1 (Release Date: 27 August 2025)
4040

4141
| CNS Version | 16.0 | 15.1 | 14.2 |
4242
| :-----: | :-----: | :------: | :------: |
@@ -47,11 +47,11 @@ For more information, Refer [Cloud Native Stack Releases](https://github.com/NVI
4747
| CRI-O | 1.33.2 | 1.32.6 | 1.31.10 |
4848
| Kubernetes | 1.33.2 | 1.32.6 | 1.31.10 |
4949
| CNI (Calico) | 3.30.2 | 3.30.2 | 3.30.2 |
50-
| NVIDIA GPU Operator | 25.3.1 | 25.3.1 | 25.3.1 |
50+
| NVIDIA GPU Operator | 25.3.2 | 25.3.2 | 25.3.2 |
5151
| NVIDIA Network Operator | N/A | 25.4.0 | 25.4.0 |
5252
| NVIDIA NIM Operator | 2.0.1 | 2.0.1 | 2.0.1 |
5353
| NVIDIA Nsight Operator | 1.1.2 | 1.1.2 | 1.1.2 |
54-
| NVIDIA Data Center Driver | 570.158.01 | 570.158.01 | 570.158.01 |
54+
| NVIDIA Data Center Driver | 580.65.06 | 580.65.06 | 580.65.06 |
5555
| Helm | 3.18.3 | 3.18.3 | 3.18.3 |
5656

5757
> NOTE: NVIDIA Network Operator is not Supported with CNS 16.0 yet

‎cns.json‎

Lines changed: 8 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -13,9 +13,9 @@
1313
"k8s version": "1.33.2",
1414
"Calico": "v3.30.2",
1515
"helm version": "3.18.3",
16-
"NVIDIA GPU Operator": "25.3.1",
16+
"NVIDIA GPU Operator": "25.3.2",
1717
"NVIDIA Network Operator": "25.4.0",
18-
"NVIDIA DataCenter Driver": "570.158.01"
18+
"NVIDIA DataCenter Driver": "580.65.06"
1919
}
2020
}
2121
]
@@ -33,9 +33,9 @@
3333
"k8s version": "1.33.2",
3434
"Calico": "v3.30.2",
3535
"helm version": "3.18.3",
36-
"NVIDIA GPU Operator": "25.3.1",
36+
"NVIDIA GPU Operator": "25.3.2",
3737
"NVIDIA Network Operator": "25.4.0",
38-
"NVIDIA DataCenter Driver": "570.158.01"
38+
"NVIDIA DataCenter Driver": "580.65.06"
3939
}
4040
}
4141
]
@@ -52,9 +52,9 @@
5252
"k8s version": "1.32.6",
5353
"Calico": "v3.30.2",
5454
"helm version": "3.18.3",
55-
"NVIDIA GPU Operator": "25.3.1",
55+
"NVIDIA GPU Operator": "25.3.2",
5656
"NVIDIA Network Operator": "25.4.0",
57-
"NVIDIA DataCenter Driver": "570.158.01"
57+
"NVIDIA DataCenter Driver": "580.65.06"
5858
}
5959
}
6060
]
@@ -90,9 +90,9 @@
9090
"k8s version": "1.31.10",
9191
"Calico": "v3.30.2",
9292
"helm version": "3.18.3",
93-
"NVIDIA GPU Operator": "25.3.1",
93+
"NVIDIA GPU Operator": "25.3.2",
9494
"NVIDIA Network Operator": "25.4.0",
95-
"NVIDIA DataCenter Driver": "570.158.01"
95+
"NVIDIA DataCenter Driver": "580.65.06"
9696
}
9797
}
9898
]

‎install-guides/Ubuntu-22-04_Server_Developer-x86-arm64_v14.2.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -10,7 +10,7 @@ NVIDIA Cloud Native Stack v14.2 includes:
1010
- Helm 3.18.3
1111
- NVIDIA GPU Driver: 570.158.01
1212
- NVIDIA Container Toolkit: 1.17.8
13-
- NVIDIA GPU Operator 25.3.1
13+
- NVIDIA GPU Operator 25.3.2
1414
- NVIDIA K8S Device Plugin: 0.17.2
1515
- NVIDIA DCGM-Exporter: 4.2.3-4.1.3
1616
- NVIDIA DCGM: 4.2.3-1
@@ -728,7 +728,7 @@ Install GPU Operator:
728728
`NOTE:` As we are preinstalled with NVIDIA Driver and NVIDIA Container Toolkit, we need to set as `false` when installing the GPU Operator
729729

730730
```
731-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator --devel nvidia/gpu-operator --set driver.enabled=false,toolkit.enabled=false --wait --generate-name
731+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator --devel nvidia/gpu-operator --set driver.enabled=false,toolkit.enabled=false --wait --generate-name
732732
```
733733

734734
#### Validating the State of the GPU Operator:
@@ -945,7 +945,7 @@ Execute the below commands to uninstall the GPU Operator:
945945
```
946946
$ helm ls
947947
NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION
948-
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.1 v25.3.1
948+
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.2 v25.3.2
949949
950950
$ helm del gpu-operator-1606173805 -n nvidia-gpu-operator
951951
```

‎install-guides/Ubuntu-22-04_Server_x86-arm64_v14.2.md‎

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,7 @@ NVIDIA Cloud Native Stack v14.2 includes:
88
- Containerd 2.1.3
99
- Kubernetes version 1.31.10
1010
- Helm 3.18.3
11-
- NVIDIA GPU Operator 25.3.1
11+
- NVIDIA GPU Operator 25.3.2
1212
- NVIDIA GPU Driver: 570.158.01
1313
- NVIDIA Container Toolkit: 1.17.8
1414
- NVIDIA K8S Device Plugin: 0.17.2
@@ -606,7 +606,7 @@ Install GPU Operator:
606606
`NOTE:` If you installed Network Operator, please skip the below command and follow the [GPU Operator with RDMA](#GPU-Operator-with-RDMA)
607607

608608
```
609-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.version=570.124.06 --wait --generate-name
609+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.version=570.124.06 --wait --generate-name
610610
```
611611

612612
#### GPU Operator with RDMA
@@ -617,15 +617,15 @@ helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator
617617
After Network Operator installation is completed, execute the below command to install the GPU Operator to load nv_peer_mem modules:
618618

619619
```
620-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true --wait --generate-name
620+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true --wait --generate-name
621621
```
622622

623623
#### GPU Operator with Host MOFED Driver and RDMA
624624

625625
If the host is already installed MOFED driver without network operator, execute the below command to install the GPU Operator to load nv_peer_mem module
626626

627627
```
628-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true,driver.rdma.useHostMofed=true --wait --generate-name
628+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true,driver.rdma.useHostMofed=true --wait --generate-name
629629
630630
```
631631

@@ -634,7 +634,7 @@ If the host is already installed MOFED driver without network operator, execute
634634
Execute the below command to enable the GPU Direct Storage Driver on GPU Operator
635635

636636
```
637-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set gds.enabled=true
637+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set gds.enabled=true
638638
```
639639
For more information refer, [GPU Direct Storage](https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/gpu-operator-rdma.html)
640640

@@ -1124,7 +1124,7 @@ Execute the below commands to uninstall the GPU Operator:
11241124
```
11251125
$ helm ls
11261126
NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION
1127-
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.1 25.3.1
1127+
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.2 25.3.2
11281128
11291129
$ helm del gpu-operator-1606173805 -n nvidia-gpu-operator
11301130

‎install-guides/Ubuntu-24-04_Server_Developer-x86-arm64_v15.1.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -10,7 +10,7 @@ NVIDIA Cloud Native Stack v15.1 includes:
1010
- Helm 3.18.3
1111
- NVIDIA GPU Driver: 570.158.01
1212
- NVIDIA Container Toolkit: 1.17.8
13-
- NVIDIA GPU Operator 25.3.1
13+
- NVIDIA GPU Operator 25.3.2
1414
- NVIDIA K8S Device Plugin: 0.17.2
1515
- NVIDIA DCGM-Exporter: 4.2.3-4.1.3
1616
- NVIDIA DCGM: 4.2.3-1
@@ -727,7 +727,7 @@ Install GPU Operator:
727727
`NOTE:` As we are preinstalled with NVIDIA Driver and NVIDIA Container Toolkit, we need to set as `false` when installing the GPU Operator
728728

729729
```
730-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator --devel nvidia/gpu-operator --set driver.enabled=false,toolkit.enabled=false --wait --generate-name
730+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator --devel nvidia/gpu-operator --set driver.enabled=false,toolkit.enabled=false --wait --generate-name
731731
```
732732

733733
#### Validating the State of the GPU Operator:
@@ -944,7 +944,7 @@ Execute the below commands to uninstall the GPU Operator:
944944
```
945945
$ helm ls
946946
NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION
947-
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.1 v25.3.1
947+
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.2 v25.3.2
948948
949949
$ helm del gpu-operator-1606173805 -n nvidia-gpu-operator
950950
```

‎install-guides/Ubuntu-24-04_Server_Developer-x86-arm64_v16.0.md‎

Lines changed: 3 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -10,7 +10,7 @@ NVIDIA Cloud Native Stack v16.0 includes:
1010
- Helm 3.18.3
1111
- NVIDIA GPU Driver: 570.158.01
1212
- NVIDIA Container Toolkit: 1.17.8
13-
- NVIDIA GPU Operator 25.3.1
13+
- NVIDIA GPU Operator 25.3.2
1414
- NVIDIA K8S Device Plugin: 0.17.2
1515
- NVIDIA DCGM-Exporter: 4.2.3-4.1.3
1616
- NVIDIA DCGM: 4.2.3-1
@@ -727,7 +727,7 @@ Install GPU Operator:
727727
`NOTE:` As we are preinstalled with NVIDIA Driver and NVIDIA Container Toolkit, we need to set as `false` when installing the GPU Operator
728728

729729
```
730-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator --devel nvidia/gpu-operator --set driver.enabled=false,toolkit.enabled=false --wait --generate-name
730+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator --devel nvidia/gpu-operator --set driver.enabled=false,toolkit.enabled=false --wait --generate-name
731731
```
732732

733733
#### Validating the State of the GPU Operator:
@@ -944,7 +944,7 @@ Execute the below commands to uninstall the GPU Operator:
944944
```
945945
$ helm ls
946946
NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION
947-
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.1 v25.3.1
947+
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.2 v25.3.2
948948
949949
$ helm del gpu-operator-1606173805 -n nvidia-gpu-operator
950950
```

‎install-guides/Ubuntu-24-04_Server_x86-arm64_v15.1.md‎

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,7 @@ NVIDIA Cloud Native Stack v15.1 includes:
88
- Containerd 2.1.3
99
- Kubernetes version 1.32.6
1010
- Helm 3.18.3
11-
- NVIDIA GPU Operator 25.3.1
11+
- NVIDIA GPU Operator 25.3.2
1212
- NVIDIA GPU Driver: 570.158.01
1313
- NVIDIA Container Toolkit: 1.17.8
1414
- NVIDIA K8S Device Plugin: 0.17.2
@@ -606,7 +606,7 @@ Install GPU Operator:
606606
`NOTE:` If you installed Network Operator, please skip the below command and follow the [GPU Operator with RDMA](#GPU-Operator-with-RDMA)
607607

608608
```
609-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.version=570.124.06 --wait --generate-name
609+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.version=570.124.06 --wait --generate-name
610610
```
611611

612612
#### GPU Operator with RDMA
@@ -617,15 +617,15 @@ helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator
617617
After Network Operator installation is completed, execute the below command to install the GPU Operator to load nv_peer_mem modules:
618618

619619
```
620-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true --wait --generate-name
620+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true --wait --generate-name
621621
```
622622

623623
#### GPU Operator with Host MOFED Driver and RDMA
624624

625625
If the host is already installed MOFED driver without network operator, execute the below command to install the GPU Operator to load nv_peer_mem module
626626

627627
```
628-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true,driver.rdma.useHostMofed=true --wait --generate-name
628+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true,driver.rdma.useHostMofed=true --wait --generate-name
629629
630630
```
631631

@@ -634,7 +634,7 @@ If the host is already installed MOFED driver without network operator, execute
634634
Execute the below command to enable the GPU Direct Storage Driver on GPU Operator
635635

636636
```
637-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set gds.enabled=true
637+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set gds.enabled=true
638638
```
639639
For more information refer, [GPU Direct Storage](https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/gpu-operator-rdma.html)
640640

@@ -1124,7 +1124,7 @@ Execute the below commands to uninstall the GPU Operator:
11241124
```
11251125
$ helm ls
11261126
NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION
1127-
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.1 25.3.1
1127+
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.2 25.3.2
11281128
11291129
$ helm del gpu-operator-1606173805 -n nvidia-gpu-operator
11301130

‎install-guides/Ubuntu-24-04_Server_x86-arm64_v16.0.md‎

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,7 @@ NVIDIA Cloud Native Stack v16.0 includes:
88
- Containerd 2.1.3
99
- Kubernetes version 1.33.2
1010
- Helm 3.18.3
11-
- NVIDIA GPU Operator 25.3.1
11+
- NVIDIA GPU Operator 25.3.2
1212
- NVIDIA GPU Driver: 570.158.01
1313
- NVIDIA Container Toolkit: 1.17.8
1414
- NVIDIA K8S Device Plugin: 0.17.2
@@ -606,7 +606,7 @@ Install GPU Operator:
606606
`NOTE:` If you installed Network Operator, please skip the below command and follow the [GPU Operator with RDMA](#GPU-Operator-with-RDMA)
607607

608608
```
609-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.version=570.124.06 --wait --generate-name
609+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.version=570.124.06 --wait --generate-name
610610
```
611611

612612
#### GPU Operator with RDMA
@@ -617,15 +617,15 @@ helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator
617617
After Network Operator installation is completed, execute the below command to install the GPU Operator to load nv_peer_mem modules:
618618

619619
```
620-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true --wait --generate-name
620+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true --wait --generate-name
621621
```
622622

623623
#### GPU Operator with Host MOFED Driver and RDMA
624624

625625
If the host is already installed MOFED driver without network operator, execute the below command to install the GPU Operator to load nv_peer_mem module
626626

627627
```
628-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true,driver.rdma.useHostMofed=true --wait --generate-name
628+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set driver.rdma.enabled=true,driver.rdma.useHostMofed=true --wait --generate-name
629629
630630
```
631631

@@ -634,7 +634,7 @@ If the host is already installed MOFED driver without network operator, execute
634634
Execute the below command to enable the GPU Direct Storage Driver on GPU Operator
635635

636636
```
637-
helm install --version 25.3.1 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set gds.enabled=true
637+
helm install --version 25.3.2 --create-namespace --namespace nvidia-gpu-operator nvidia/gpu-operator --set gds.enabled=true
638638
```
639639
For more information refer, [GPU Direct Storage](https://docs.nvidia.com/datacenter/cloud-native/gpu-operator/gpu-operator-rdma.html)
640640

@@ -1124,7 +1124,7 @@ Execute the below commands to uninstall the GPU Operator:
11241124
```
11251125
$ helm ls
11261126
NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION
1127-
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.1 25.3.1
1127+
gpu-operator-1606173805 nvidia-gpu-operator 1 2025-03-31 20:23:28.063421701 +0000 UTC deployed gpu-operator-25.3.2 25.3.2
11281128
11291129
$ helm del gpu-operator-1606173805 -n nvidia-gpu-operator
11301130

‎playbooks/cns-uninstall.yaml‎

Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -145,6 +145,16 @@
145145
failed_when: false
146146
no_log: True
147147

148+
- name: Reset iptables default policies to ACCEPT
149+
ansible.builtin.shell: "{{ item }}"
150+
become: true
151+
with_items:
152+
- iptables -P INPUT ACCEPT
153+
- iptables -P FORWARD ACCEPT
154+
- iptables -P OUTPUT ACCEPT
155+
ignore_errors: yes
156+
failed_when: false
157+
148158
- name: IPTables Cleanup
149159
ignore_errors: yes
150160
failed_when: false

0 commit comments

Comments
 (0)