Merge branch 'main' into ibmcloud-gh-runners

Author: deekay2310

.github/workflows/build-go.yaml

@@ -22,7 +22,8 @@ jobs:
         rm_cmd: "rmz"
 
     - name: Checkout code
-      uses: actions/checkout@v6
+
+      uses: actions/checkout@v7
 
     - name: Set up Go
       uses: actions/setup-go@v6

.github/workflows/build-img-ghrunner-test.yaml

@@ -21,7 +21,7 @@ jobs:
         rm_cmd: "rmz"
 
     - name: Checkout code
-      uses: actions/checkout@v6
+      uses: actions/checkout@v7
 
     - name: Build image for PR
       if: ${{ github.event_name == 'pull_request' }}

.github/workflows/build-oci.yaml

@@ -40,7 +40,7 @@ jobs:
           remove_haskell: true
           rm_cmd: "rmz"
       - name: Checkout code
-        uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
       - name: Install Podman (only for arm64)
         if: contains(matrix.os, 'arm')
         run: |
@@ -121,7 +121,7 @@ jobs:
     runs-on: ubuntu-24.04
     steps:
       - name: Checkout code
-        uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
+        uses: actions/checkout@9c091bb21b7c1c1d1991bb908d89e4e9dddfe3e0 # v7.0.0
       - name: Set version from tag
         if: startsWith(github.ref, 'refs/tags/')
         run: echo "VERSION=${GITHUB_REF_NAME#v}" >> $GITHUB_ENV

.github/workflows/build-on-hosted-runner.yaml

@@ -23,7 +23,7 @@ jobs:
     needs: hosted_runner_provision
     steps:
       - name: Code checkout
-        uses: actions/checkout@v6
+        uses: actions/checkout@v7
       - name: Set up Go
         uses: actions/setup-go@v6
         with:

.github/workflows/tkn-bundle.yaml

@@ -14,7 +14,7 @@ jobs:
     runs-on: ubuntu-24.04
     steps:
     - name: Checkout code
-      uses: actions/checkout@v6
+      uses: actions/checkout@v7
 
     - name: Create k8s Kind Cluster
       uses: helm/kind-action@v1
@@ -33,7 +33,7 @@ jobs:
     runs-on: ubuntu-24.04
     steps:
     - name: Checkout code
-      uses: actions/checkout@v6
+      uses: actions/checkout@v7
 
     - name: Log in to quay.io Registry
       uses: redhat-actions/podman-login@v1

Makefile

@@ -9,9 +9,9 @@ TKN_IMG ?= quay.io/redhat-developer/mapt:v${VERSION}-tkn
 # renovate: datasource=github-releases depName=cirruslabs/cirrus-cli
 CIRRUS_CLI ?= v1.0.0
 # renovate: datasource=github-releases depName=actions/runner
-GITHUB_RUNNER ?= 2.334.0
+GITHUB_RUNNER ?= 2.335.1
 # renovate: datasource=gitlab-releases depName=gitlab-org/gitlab-runner
-GITLAB_RUNNER ?= 19.0.1
+GITLAB_RUNNER ?= 19.1.0
 # renovate: datasource=github-releases depName=open-telemetry/opentelemetry-collector-releases
 OTELCOL_VERSION ?= 0.151.0
 

cmd/mapt/cmd/aws/hosts/rhelai.go

@@ -1,6 +1,8 @@
 package hosts
 
 import (
+	"fmt"
+
 	awsParams "github.com/redhat-developer/mapt/cmd/mapt/cmd/aws/params"
 	"github.com/redhat-developer/mapt/cmd/mapt/cmd/params"
 	maptContext "github.com/redhat-developer/mapt/pkg/manager/context"
@@ -14,6 +16,9 @@ import (
 const (
 	cmdRHELAI     = "rhel-ai"
 	cmdRHELAIDesc = "manage rhel ai host"
+
+	cmdRHELAIListVersions     = "list-versions"
+	cmdRHELAIListVersionsDesc = "list available RHEL AI versions"
 )
 
 func GetRHELAICmd() *cobra.Command {
@@ -32,7 +37,7 @@ func GetRHELAICmd() *cobra.Command {
 	params.AddCommonFlags(flagSet)
 	c.PersistentFlags().AddFlagSet(flagSet)
 
-	c.AddCommand(getRHELAICreate(), getRHELAIDestroy())
+	c.AddCommand(getRHELAICreate(), getRHELAIDestroy(), getRHELAIListVersions())
 	return c
 }
 
@@ -107,3 +112,27 @@ func getRHELAIDestroy() *cobra.Command {
 	c.PersistentFlags().AddFlagSet(flagSet)
 	return c
 }
+
+func getRHELAIListVersions() *cobra.Command {
+	c := &cobra.Command{
+		Use:   cmdRHELAIListVersions,
+		Short: cmdRHELAIListVersionsDesc,
+		RunE: func(cmd *cobra.Command, args []string) error {
+			if err := viper.BindPFlags(cmd.Flags()); err != nil {
+				return err
+			}
+			versions, err := rhelai.ListVersions(cmd.Context(), viper.GetString(params.RhelAIAccelerator))
+			if err != nil {
+				return err
+			}
+			for _, v := range versions {
+				fmt.Println(v)
+			}
+			return nil
+		},
+	}
+	flagSet := pflag.NewFlagSet(cmdRHELAIListVersions, pflag.ExitOnError)
+	flagSet.StringP(params.RhelAIAccelerator, "", params.RhelAIAccelearatorDefault, params.RhelAIAccelearatorDesc)
+	c.PersistentFlags().AddFlagSet(flagSet)
+	return c
+}

cmd/mapt/cmd/azure/hosts/rhelai.go

@@ -1,6 +1,8 @@
 package hosts
 
 import (
+	"fmt"
+
 	"github.com/redhat-developer/mapt/cmd/mapt/cmd/params"
 	maptContext "github.com/redhat-developer/mapt/pkg/manager/context"
 	rhelai "github.com/redhat-developer/mapt/pkg/provider/azure/action/rhel-ai"
@@ -13,6 +15,9 @@ import (
 const (
 	cmdRHELAI     = "rhel-ai"
 	cmdRHELAIDesc = "manage rhel ai host"
+
+	cmdRHELAIListVersions     = "list-versions"
+	cmdRHELAIListVersionsDesc = "list available RHEL AI versions in the Azure Compute Gallery"
 )
 
 func GetRHELAICmd() *cobra.Command {
@@ -31,7 +36,7 @@ func GetRHELAICmd() *cobra.Command {
 	params.AddCommonFlags(flagSet)
 	c.PersistentFlags().AddFlagSet(flagSet)
 
-	c.AddCommand(getRHELAICreate(), getRHELAIDestroy())
+	c.AddCommand(getRHELAICreate(), getRHELAIDestroy(), getRHELAIListVersions())
 	return c
 }
 
@@ -58,6 +63,7 @@ func getRHELAICreate() *cobra.Command {
 					Version:        viper.GetString(params.RhelAIVersion),
 					Accelerator:    viper.GetString(params.RhelAIAccelerator),
 					CustomImage:    viper.GetString(params.RhelAICustomImage),
+					Marketplace:    viper.GetBool(params.RhelAIMarketplace),
 					ComputeRequest: params.ComputeRequestArgs(),
 					Spot:           params.SpotArgs(),
 					Timeout:        viper.GetString(params.Timeout),
@@ -70,6 +76,7 @@ func getRHELAICreate() *cobra.Command {
 	flagSet.StringP(params.RhelAIVersion, "", params.RhelAIVersionDefault, params.RhelAIVersionDesc)
 	flagSet.StringP(params.RhelAIAccelerator, "", params.RhelAIAccelearatorDefault, params.RhelAIAccelearatorDesc)
 	flagSet.StringP(params.RhelAICustomImage, "", "", params.RhelAICustomImageDesc)
+	flagSet.Bool(params.RhelAIMarketplace, false, params.RhelAIMarketplaceDesc)
 	flagSet.StringP(params.Timeout, "", "", params.TimeoutDesc)
 	params.AddComputeRequestFlags(flagSet)
 	params.AddSpotFlags(flagSet)
@@ -104,3 +111,27 @@ func getRHELAIDestroy() *cobra.Command {
 	c.PersistentFlags().AddFlagSet(flagSet)
 	return c
 }
+
+func getRHELAIListVersions() *cobra.Command {
+	c := &cobra.Command{
+		Use:   cmdRHELAIListVersions,
+		Short: cmdRHELAIListVersionsDesc,
+		RunE: func(cmd *cobra.Command, args []string) error {
+			if err := viper.BindPFlags(cmd.Flags()); err != nil {
+				return err
+			}
+			versions, err := rhelai.ListVersions(cmd.Context(), viper.GetString(params.RhelAIAccelerator))
+			if err != nil {
+				return err
+			}
+			for _, v := range versions {
+				fmt.Println(v)
+			}
+			return nil
+		},
+	}
+	flagSet := pflag.NewFlagSet(cmdRHELAIListVersions, pflag.ExitOnError)
+	flagSet.StringP(params.RhelAIAccelerator, "", params.RhelAIAccelearatorDefault, params.RhelAIAccelearatorDesc)
+	c.PersistentFlags().AddFlagSet(flagSet)
+	return c
+}

cmd/mapt/cmd/params/params.go

@@ -128,6 +128,8 @@ const (
 	RhelAIAccelearatorDefault string = "cuda"
 	RhelAICustomImage         string = "custom-image"
 	RhelAICustomImageDesc     string = "custom image name to spin RHEL AI OS (AMI name for AWS, image name for Azure)"
+	RhelAIMarketplace         string = "marketplace"
+	RhelAIMarketplaceDesc     string = "use the cloud provider's marketplace RHEL AI image instead of a shared gallery or custom image"
 
 	// Serverless
 	Timeout        string = "timeout"

docs/aws/rhelai.md

@@ -0,0 +1,110 @@
+# Overview
+
+mapt offers operations to manage RHEL AI environments on AWS. RHEL AI instances are GPU-enabled machines with pre-installed RHEL AI images, suitable for AI/ML workloads.
+
+## Operations
+
+### List Versions
+
+List available RHEL AI versions for a given accelerator type:
+
+```bash
+mapt aws rhel-ai list-versions -h
+list-versions
+
+Usage:
+  mapt aws rhel-ai list-versions [flags]
+
+Flags:
+      --accelerator string   accelerator type. Valid types: cuda and rocm (default "cuda")
+  -h, --help                 help for list-versions
+
+Global Flags:
+      --backed-url string     backed for stack state. Can be a local path with format file:///path/subpath or s3 s3://existing-bucket
+      --project-name string   project name to identify the instance of the stack
+```
+
+#### Container
+
+```bash
+podman run -it --rm \
+    -e AWS_ACCESS_KEY_ID=XXX \
+    -e AWS_SECRET_ACCESS_KEY=XXX \
+    -e AWS_DEFAULT_REGION=us-east-1 \
+    quay.io/redhat-developer/mapt:0.7.0-dev aws \
+        rhel-ai list-versions \
+        --accelerator cuda
+```
+
+### Create
+
+This will create a RHEL AI instance according to params specified:
+
+```bash
+mapt aws rhel-ai create -h
+create
+
+Usage:
+  mapt aws rhel-ai create [flags]
+
+Flags:
+      --accelerator string         accelerator type. Valid types: cuda and rocm (default "cuda")
+      --conn-details-output string path to export host connection information (host, username and privateKey)
+      --cpus int32                 Number of CPUs for the cloud instance (default 8)
+      --custom-image string        custom AMI name (overrides version and accelerator)
+      --disk-size int              Disk size in GB (default 2000)
+      --gpus int32                 Number of GPUs
+      --memory int32               Amount of RAM for the cloud instance in GiB (default 64)
+      --spot                       if spot is set the spot prices across all regions will be checked and machine will be started on best spot option (price / eviction)
+      --spot-eviction-tolerance string  if spot is enabled we can define the minimum tolerance level of eviction. Allowed values are: lowest, low, medium, high or highest (default "lowest")
+      --tags stringToString        tags to add on each resource (--tags name1=value1,name2=value2) (default [])
+      --timeout string             set a timeout for the instance (e.g. 4h)
+      --version string             version for the RHELAI OS (default "3.0.0")
+  -h, --help                       help for create
+
+Global Flags:
+      --backed-url string     backed for stack state. Can be a local path with format file:///path/subpath or s3 s3://existing-bucket
+      --project-name string   project name to identify the instance of the stack
+```
+
+#### Outputs
+
+It will create a RHEL AI instance and will give as result several files located at path defined by `--conn-details-output`:
+
+* **host**: host for the instance (load balancer DNS if spot)
+* **username**: username to connect to the machine
+* **id_rsa**: private key to connect to the machine
+
+Also, it will create a state folder holding the state for the created resources at AWS, the path for this folder is defined within `--backed-url`, the content from that folder is required with the same project name (`--project-name`) in order to destroy the resources.
+
+#### Container
+
+When running the container image it is required to pass the authentication information as variables, following a sample snippet on how to create an instance with default values:
+
+```bash
+podman run -d --name mapt-rhelai \
+    -v ${PWD}:/workspace:z \
+    -e AWS_ACCESS_KEY_ID=XXX \
+    -e AWS_SECRET_ACCESS_KEY=XXX \
+    -e AWS_DEFAULT_REGION=us-east-1 \
+    quay.io/redhat-developer/mapt:0.7.0-dev aws \
+        rhel-ai create \
+        --project-name mapt-rhelai \
+        --backed-url file:///workspace \
+        --conn-details-output /workspace \
+        --spot
+```
+
+### Destroy
+
+```bash
+podman run -d --rm \
+    -v ${PWD}:/workspace:z \
+    -e AWS_ACCESS_KEY_ID=XXX \
+    -e AWS_SECRET_ACCESS_KEY=XXX \
+    -e AWS_DEFAULT_REGION=us-east-1 \
+    quay.io/redhat-developer/mapt:0.7.0-dev aws \
+        rhel-ai destroy \
+        --project-name mapt-rhelai \
+        --backed-url file:///workspace
+```