AISTACK_BASIC test

os-autoinst · Nov 17, 2024 · 8863197 · 8863197
1 parent 7732689
commit 8863197
Show file tree

Hide file tree

Showing 5 changed files with 648 additions and 4 deletions.
diff --git a/data/aistack/aistack_gpu_values.yaml b/data/aistack/aistack_gpu_values.yaml
@@ -0,0 +1,193 @@
+global:
+  tls:
+    # options: suse-private-ai, letsEncrypt, secret
+    source: suse-private-ai
+    issuerName: suse-private-ai
+
+    # This section to be filled out when using letsEncrypt as the tls source
+    letsEncrypt:
+      environment: staging
+      email: [email protected]
+      ingress:
+        class: ""
+ollama:
+  ollama:
+    gpu:
+     enabled: true
+     type: 'nvidia'
+     number: 1
+  ingress:
+    enabled: false
+open-webui:
+  ollamaUrls:
+  - http://suse-private-ai-ollama.suse-private-ai.svc.cluster.local:11434
+  persistence:
+    enabled: true
+    storageClass: local-path
+  ollama:
+    enabled: false
+  pipelines:
+    enabled: false
+    persistence:
+      storageClass: local-path
+  ingress:
+    enabled: true
+    class: ""
+    annotations:
+      nginx.ingress.kubernetes.io/ssl-redirect: "true"
+    host: suse-ollama-webui
+    tls: true
+    existingSecret: suse-private-ai-tls
+  extraEnvVars:
+  - name: DEFAULT_MODELS
+    value: "gemma:2b"
+  - name: DEFAULT_USER_ROLE
+    value: "user"
+  - name: WEBUI_NAME
+    value: "SUSE AI"
+  - name: GLOBAL_LOG_LEVEL
+    value: INFO
+  - name: VECTOR_DB
+    value: "milvus"
+  - name: MILVUS_URI
+    value: http://suse-private-ai-milvus.suse-private-ai.svc.cluster.local:19530
+milvus:
+  enabled: True
+  cluster:
+    enabled: True
+  standalone:
+    persistence:
+      persistentVolumeClaim:
+        storageClass: local-path
+  etcd:
+    replicaCount: 1
+    persistence:
+      storageClassName: local-path
+  minio:
+    mode: distributed
+    replicas: 4
+    rootUser: "admin"
+    rootPassword: "adminminio"
+    persistence:
+      storageClass: local-path
+    resources:
+      requests:
+        memory: 1024Mi
+  pulsar:
+    enabled: True
+    affinity:
+      anti_affinity: false
+    autorecovery:
+      resources:
+        requests:
+          cpu: 0.1
+          memory: 256Mi
+    proxy:
+      replicaCount: 1
+      resources:
+        requests:
+          cpu: 0.2
+          memory: 256Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms256m -Xmx256m
+        PULSAR_GC: >
+           -XX:MaxDirectMemorySize=256m
+    bookkeeper:
+      replicaCount: 2
+      volumes:
+        journal:
+          local_storage: false
+          storageClassName: local-path
+        ledgers:
+          local_storage: false
+          storageClassName: local-path
+        common:
+          local_storage: false
+          storageClassName: local-path
+      resources:
+        requests:
+          cpu: 0.2
+          memory: 512Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms512m -Xmx512m
+        PULSAR_GC: >
+          -XX:MaxDirectMemorySize=512m
+          -Dio.netty.leakDetectionLevel=disabled
+          -Dio.netty.recycler.linkCapacity=1024
+          -XX:+UseG1GC -XX:MaxGCPauseMillis=10
+          -XX:+ParallelRefProcEnabled
+          -XX:+UnlockExperimentalVMOptions
+          -XX:+DoEscapeAnalysis -XX:ParallelGCThreads=32
+          -XX:ConcGCThreads=32 -XX:G1NewSizePercent=50
+          -XX:+DisableExplicitGC
+          -XX:-ResizePLAB
+          -XX:+ExitOnOutOfMemoryError
+          -XX:+PerfDisableSharedMem
+          -XX:+PrintGCDetails
+    zookeeper:
+      replicaCount: 1
+      volumes:
+        data:
+          storageClassName: local-path
+      resources:
+        requests:
+          cpu: 0.1
+          memory: 256Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms256m
+          -Xmx256m
+        PULSAR_GC: >
+          -Dcom.sun.management.jmxremote
+          -Djute.maxbuffer=10485760
+          -XX:+ParallelRefProcEnabled
+          -XX:+UnlockExperimentalVMOptions
+          -XX:+DoEscapeAnalysis -XX:+DisableExplicitGC
+          -XX:+PerfDisableSharedMem
+          -Dzookeeper.forceSync=no
+    broker:
+      replicaCount: 2
+      volumes:
+        data:
+          storageClassName: local-path
+      resources:
+        requests:
+          cpu: 0.2
+          memory: 512Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms512m
+          -Xmx512m
+        PULSAR_GC: >
+          -XX:MaxDirectMemorySize=512m
+          -Dio.netty.leakDetectionLevel=disabled
+          -Dio.netty.recycler.linkCapacity=1024
+          -XX:+ParallelRefProcEnabled
+          -XX:+UnlockExperimentalVMOptions
+          -XX:+DoEscapeAnalysis
+          -XX:ParallelGCThreads=32
+          -XX:ConcGCThreads=32
+          -XX:G1NewSizePercent=50
+          -XX:+DisableExplicitGC
+          -XX:-ResizePLAB
+          -XX:+ExitOnOutOfMemoryError
+    autorecovery:
+      resources:
+        requests:
+          memory: 512Mi
+          cpu: 1
+#  #indexNode:
+  #  resources:
+  #    requests:
+  #      nvidia.com/gpu: "1"
+  #    limits:
+  #      nvidia.com/gpu: "1"
+  #queryNode:
+  #  resources:
+  #    requests:
+  #      nvidia.com/gpu: "1"
+  #    limits:
+  #      nvidia.com/gpu: "1"
+#
diff --git a/data/aistack/aistack_values.yaml b/data/aistack/aistack_values.yaml
@@ -0,0 +1,192 @@
+global:
+  tls:
+    # options: suse-private-ai, letsEncrypt, secret
+    source: suse-private-ai
+    issuerName: suse-private-ai
+
+    # This section to be filled out when using letsEncrypt as the tls source
+    letsEncrypt:
+      environment: staging
+      email: [email protected]
+      ingress:
+        class: "nginx"
+ollama:
+  ollama:
+    gpu:
+     enabled: false
+  ingress:
+    enabled: false
+open-webui:
+  ollamaUrls:
+  - http://suse-private-ai-ollama.suse-private-ai.svc.cluster.local:11434
+  persistence:
+    enabled: true
+    storageClass: local-path
+  ollama:
+    enabled: false
+  pipelines:
+    enabled: false
+    persistence:
+      storageClass: local-path
+  ingress:
+    enabled: true
+    class: ""
+    annotations:
+      nginx.ingress.kubernetes.io/ssl-redirect: "true"
+      nginx.ingress.kubernetes.io/ssl-services: "open-webui"
+    host: suse-ollama-webui
+    tls: true
+    existingSecret: suse-private-ai-tls
+  extraEnvVars:
+  - name: DEFAULT_MODELS
+    value: "gemma:2b"
+  - name: DEFAULT_USER_ROLE
+    value: "user"
+  - name: WEBUI_NAME
+    value: "SUSE AI"
+  - name: GLOBAL_LOG_LEVEL
+    value: INFO
+  - name: VECTOR_DB
+    value: "milvus"
+  - name: MILVUS_URI
+    value: http://suse-private-ai-milvus.suse-private-ai.svc.cluster.local:19530
+milvus:
+  enabled: True
+  cluster:
+    enabled: True
+  standalone:
+    persistence:
+      persistentVolumeClaim:
+        storageClass: local-path
+  etcd:
+    replicaCount: 1
+    persistence:
+      storageClassName: local-path
+  minio:
+    mode: distributed
+    replicas: 4
+    rootUser: "admin"
+    rootPassword: "adminminio"
+    persistence:
+      storageClass: local-path
+    resources:
+      requests:
+        memory: 1024Mi
+  pulsar:
+    enabled: True
+    affinity:
+      anti_affinity: false
+    autorecovery:
+      resources:
+        requests:
+          cpu: 0.1
+          memory: 256Mi
+    proxy:
+      replicaCount: 1
+      resources:
+        requests:
+          cpu: 0.2
+          memory: 256Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms256m -Xmx256m
+        PULSAR_GC: >
+           -XX:MaxDirectMemorySize=256m
+    bookkeeper:
+      replicaCount: 2
+      volumes:
+        journal:
+          local_storage: false
+          storageClassName: local-path
+        ledgers:
+          local_storage: false
+          storageClassName: local-path
+        common:
+          local_storage: false
+          storageClassName: local-path
+      resources:
+        requests:
+          cpu: 0.2
+          memory: 512Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms512m -Xmx512m
+        PULSAR_GC: >
+          -XX:MaxDirectMemorySize=512m
+          -Dio.netty.leakDetectionLevel=disabled
+          -Dio.netty.recycler.linkCapacity=1024
+          -XX:+UseG1GC -XX:MaxGCPauseMillis=10
+          -XX:+ParallelRefProcEnabled
+          -XX:+UnlockExperimentalVMOptions
+          -XX:+DoEscapeAnalysis -XX:ParallelGCThreads=32
+          -XX:ConcGCThreads=32 -XX:G1NewSizePercent=50
+          -XX:+DisableExplicitGC
+          -XX:-ResizePLAB
+          -XX:+ExitOnOutOfMemoryError
+          -XX:+PerfDisableSharedMem
+          -XX:+PrintGCDetails
+    zookeeper:
+      replicaCount: 1
+      volumes:
+        data:
+          storageClassName: local-path
+      resources:
+        requests:
+          cpu: 0.1
+          memory: 256Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms256m
+          -Xmx256m
+        PULSAR_GC: >
+          -Dcom.sun.management.jmxremote
+          -Djute.maxbuffer=10485760
+          -XX:+ParallelRefProcEnabled
+          -XX:+UnlockExperimentalVMOptions
+          -XX:+DoEscapeAnalysis -XX:+DisableExplicitGC
+          -XX:+PerfDisableSharedMem
+          -Dzookeeper.forceSync=no
+    broker:
+      replicaCount: 2
+      volumes:
+        data:
+          storageClassName: local-path
+      resources:
+        requests:
+          cpu: 0.2
+          memory: 512Mi
+      configData:
+        PULSAR_MEM: >
+          -Xms512m
+          -Xmx512m
+        PULSAR_GC: >
+          -XX:MaxDirectMemorySize=512m
+          -Dio.netty.leakDetectionLevel=disabled
+          -Dio.netty.recycler.linkCapacity=1024
+          -XX:+ParallelRefProcEnabled
+          -XX:+UnlockExperimentalVMOptions
+          -XX:+DoEscapeAnalysis
+          -XX:ParallelGCThreads=32
+          -XX:ConcGCThreads=32
+          -XX:G1NewSizePercent=50
+          -XX:+DisableExplicitGC
+          -XX:-ResizePLAB
+          -XX:+ExitOnOutOfMemoryError
+    autorecovery:
+      resources:
+        requests:
+          memory: 512Mi
+          cpu: 1
+#  #indexNode:
+  #  resources:
+  #    requests:
+  #      nvidia.com/gpu: "1"
+  #    limits:
+  #      nvidia.com/gpu: "1"
+  #queryNode:
+  #  resources:
+  #    requests:
+  #      nvidia.com/gpu: "1"
+  #    limits:
+  #      nvidia.com/gpu: "1"
+#
diff --git a/data/aistack/nvidia_gpu_values.yaml b/data/aistack/nvidia_gpu_values.yaml
@@ -0,0 +1,10 @@
+toolkit:
+  env:
+  - name: CONTAINERD_CONFIG
+    value: /var/lib/rancher/rke2/agent/etc/containerd/config.toml.tmpl
+  - name: CONTAINERD_SOCKET
+    value: /run/k3s/containerd/containerd.sock
+  - name: CONTAINERD_RUNTIME_CLASS
+    value: nvidia
+  - name: CONTAINERD_SET_AS_DEFAULT
+    value: "true"