fix: update tsdb installation, pod exec role, gpu model info (#77)

Author: Code2Life

README.md

@@ -49,29 +49,11 @@ WIP
 - [Getting Started on VM](https://tensor-fusion.ai/guide/deployment-vm)
 - [Deploy Self-hosted Community Edition](https://tensor-fusion.ai/guide/self-host)
 
-### Try it out
+<!-- (TODO: Asciinema) -->
 
-- Explore the demo account: [Demo Console - Working in progress](https://app.tensor-fusion.ai?hint=demo)
+<!-- ### Playground
 
-- Run following command to try TensorFusion in 3 minutes
-
-```bash
-# Step 1: Install TensorFusion in Kubernetes
-helm install --repo https://nexusgpu.github.io/tensor-fusion/ --create-namespace
-
-# Step 2. Onboard GPU nodes into TensorFusion cluster
-kubectl apply -f https://raw.githubusercontent.com/NexusGPU/tensor-fusion/main/manifests/gpu-node.yaml
-
-# Step 3. Check if cluster and pool is ready
-kubectl get gpupools -o wide && kubectl get gpunodes -o wide
-
-# Step3. Create an inference app using virtual, remote GPU resources in TensorFusion cluster
-kubectl apply -f https://raw.githubusercontent.com/NexusGPU/tensor-fusion/main/manifests/inference-app.yaml
-
-# Then you can forward the port to test inference, or exec shell
-```
-
-(TODO: Asciinema)
+- Explore the demo account: [Demo Console - Working in progress](https://app.tensor-fusion.ai?hint=demo) -->
 
 ### 💬 Discussion
 
@@ -87,28 +69,29 @@ kubectl apply -f https://raw.githubusercontent.com/NexusGPU/tensor-fusion/main/m
 ### Core GPU Virtualization Features
 
 - [x] Fractional GPU and flexible oversubscription
-- [x] GPU-over-IP, remote GPU sharing with less than 4% performance loss
-- [x] GPU VRAM expansion or swap to host RAM
+- [x] Remote GPU sharing with SOTA GPU-over-IP technology, less than 4% performance loss
+- [x] GPU VRAM expansion and hot/warm/cold tiering
 - [ ] None NVIDIA GPU/NPU vendor support
 
 ### Pooling & Scheduling & Management
 
 - [x] GPU/NPU pool management in Kubernetes
-- [x] GPU-first resource scheduler based on virtual TFlops/VRAM capacity
-- [x] GPU-first auto provisioning and bin-packing
+- [x] GPU-first scheduling and allocation, with single TFlops/MB precision
+- [x] GPU node auto provisioning/termination
+- [x] GPU compaction/bin-packing
 - [x] Seamless onboarding experience for Pytorch, TensorFlow, llama.cpp, vLLM, Tensor-RT, SGlang and all popular AI training/serving frameworks
-- [x] Basic management console and dashboards
-- [ ] Basic autoscaling policies, auto set requests/limits/replicas
-- [ ] GPU Group scheduling for LLMs
+- [x] Centralized Dashboard & Control Plane
+- [ ] GPU-first autoscaling policies, auto set requests/limits/replicas
+- [ ] Request multiple vGPUs with group scheduling for large models
 - [ ] Support different QoS levels
 
 ### Enterprise Features
 
-- [x] GPU live-migration, fastest in the world
-- [ ] Preloading and P2P distribution of container images, AI models, GPU snapshots etc.
+- [x] GPU live-migration, snapshot/distribute/restore GPU context cross cluster, fastest in the world
+- [ ] AI model registry and preloading, build your own private MaaS(Model-as-a-Service)
 - [ ] Advanced auto-scaling policies, scale to zero, rebalance of hot GPUs
 - [ ] Advanced observability features, detailed metrics & tracing/profiling of CUDA calls
-- [ ] Multi-tenancy billing based on actual usage
+- [ ] Monetization your GPU cluster by multi-tenancy usage measurement & billing report
 - [ ] Enterprise level high availability and resilience, support topology aware scheduling, GPU node auto failover etc.
 - [ ] Enterprise level security, complete on-premise deployment support, encryption in-transit & at-rest
 - [ ] Enterprise level compliance, SSO/SAML support, advanced audit, ReBAC control, SOC2 and other compliance reports available

charts/tensor-fusion/Chart.yaml

@@ -15,7 +15,7 @@ type: application
 # This is the chart version. This version number should be incremented each time you make changes
 # to the chart and its templates, including the app version.
 # Versions are expected to follow Semantic Versioning (https://semver.org/)
-version: 1.2.8
+version: 1.2.9
 
 # This is the version number of the application being deployed. This version number should be
 # incremented each time you make changes to the application. Versions are not expected to

charts/tensor-fusion/templates/gpu-public-gpu-info.yaml

@@ -29,32 +29,32 @@ data:
 
     # Ampere Architecture Series
     - model: A100_SXM4
-      fullModelName: "A100 SXM4"
+      fullModelName: "NVIDIA A100 SXM4"
       vendor: NVIDIA
       costPerHour: 1.89
       fp16TFlops: 312
     
     - model: A100_PCIe
-      fullModelName: "A100 PCIe"
+      fullModelName: "NVIDIA A100 PCIe"
       vendor: NVIDIA
       costPerHour: 1.64
       fp16TFlops: 312
 
     - model: A10
-      fullModelName: "A10"
+      fullModelName: "NVIDIA A10"
       vendor: NVIDIA
       costPerHour: 0.9
       fp16TFlops: 125
 
     # A10G has less CUDA core than A10, but with RT cores for rendering case
     - model: A10G
-      fullModelName: "A10G"
+      fullModelName: "NVIDIA A10G"
       vendor: NVIDIA
       costPerHour: 0.75 # from lambda labs
-      fp16TFlops: 125
+      fp16TFlops: 63
 
     - model: A40
-      fullModelName: "A40"
+      fullModelName: "NVIDIA A40"
       vendor: NVIDIA
       costPerHour: 0.44
       fp16TFlops: 125
@@ -67,22 +67,22 @@ data:
 
     # Ada Lovelace Architecture Series  
     - model: L4
-      fullModelName: "L4"
+      fullModelName: "NVIDIA L4"
       vendor: NVIDIA
       costPerHour: 0.43
       fp16TFlops: 121
 
     - model: L40
-      fullModelName: "L40"
+      fullModelName: "NVIDIA L40"
       vendor: NVIDIA
       costPerHour: 0.86 # should be a bit cheaper than L40s
-      fp16TFlops: 362
+      fp16TFlops: 181
 
     - model: L40s
-      fullModelName: "L40s"
+      fullModelName: "NVIDIA L40s"
       vendor: NVIDIA
       costPerHour: 0.86
-      fp16TFlops: 362
+      fp16TFlops: 181
 
     - model: RTX4090
       fullModelName: "RTX4090"
@@ -92,20 +92,20 @@ data:
 
     # Hopper Architecture Series
     - model: H100_SXM4
-      fullModelName: "H100 SXM4"
+      fullModelName: "NVIDIA H100 SXM4"
       vendor: NVIDIA
       costPerHour: 2.99
       fp16TFlops: 989
     
     - model: H100_PCIe
-      fullModelName: "H100 PCIe"
+      fullModelName: "NVIDIA H100 PCIe"
       vendor: NVIDIA
       costPerHour: 2.39
       fp16TFlops: 835
    
     # Blackwell Architecture Series
     - model: B200_SXM4
-      fullModelName: "B200 SXM4"
+      fullModelName: "NVIDIA B200 SXM4"
       vendor: NVIDIA
       costPerHour: 10.99 # unknown price,on-request
       fp16TFlops: 2250

charts/tensor-fusion/templates/greptime-standalone.yaml

@@ -1,60 +1,172 @@
 {{- if .Values.greptime.installStandalone }}
-# NOTICE: make sure greptimedb operator had been installed in your test cluster
+# NOTICE: make sure greptimedb operator had been installed in your cluster if not enable 'installStandalone'
 # cloud mode is recommended to reduce the maintenance effort
 # ```bash
 # helm repo add greptime https://greptimeteam.github.io/helm-charts/
 # helm repo update
 # helm install greptimedb-operator greptime/greptimedb-operator -n greptimedb --create-namespace
 # ```
-apiVersion: greptime.io/v1alpha1
-kind: GreptimeDBStandalone
+---
+apiVersion: v1
+kind: Namespace
 metadata:
   name: greptimedb
+---
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: greptimedb-standalone
+  namespace: greptimedb
+data:
+  config.toml: |
+    [logging]
+      dir = "/data/greptimedb/logs"
+      level = "info"
+      log_format = "text"
+
+    [storage]
+      data_home = "/data/greptimedb"
+
+    [wal]
+      dir = "/data/greptimedb/wal"
+---
+apiVersion: v1
+kind: Service
+metadata:
+  name: greptimedb-standalone
+  namespace: greptimedb
+  labels:
+    app.greptime.io/component: greptimedb-standalone
+spec:
+  selector:
+    app.greptime.io/component: greptimedb-standalone
+  ports:
+    - name: grpc
+      port: 4001
+      targetPort: 4001
+    - name: http
+      port: 4000
+      targetPort: 4000
+    - name: mysql
+      port: 4002
+      targetPort: 4002
+    - name: postgres
+      port: 4003
+      targetPort: 4003
+---
+apiVersion: apps/v1
+kind: StatefulSet
+metadata:
+  name: greptimedb-standalone
   namespace: greptimedb
+  labels:
+    app.greptime.io/component: greptimedb-standalone
 spec:
-  base:
-    main:
-      image: docker.io/greptime/greptimedb:latest
-      livenessProbe:
-        failureThreshold: 10
-        httpGet:
-          path: /health
-          port: 4000
-        periodSeconds: 5
-      readinessProbe:
-        failureThreshold: 10
-        httpGet:
-          path: /health
-          port: 4000
-        periodSeconds: 5
-      resources: {}
-      startupProbe:
-        failureThreshold: 60
-        httpGet:
-          path: /health
-          port: 4000
-        periodSeconds: 5
-  datanodeStorage:
-    dataHome: /data/greptimedb
-    fs:
-      mountPath: /data/greptimedb
-      name: datanode
-      storageRetainPolicy: Retain
-      storageSize: 20Gi
-  httpPort: 4000
-  logging:
-    format: text
-    level: info
-    logsDir: /data/greptimedb/logs
-    onlyLogToStdout: false
-    persistentWithData: false
-  mysqlPort: 4002
-  postgreSQLPort: 4003
-  rollingUpdate:
-    maxUnavailable: 1
-    partition: 0
-  rpcPort: 4001
-  service:
-    type: ClusterIP
-  version: latest
+  replicas: 1
+  selector:
+    matchLabels:
+      app.greptime.io/component: greptimedb-standalone
+  template:
+    metadata:
+      labels:
+        app.greptime.io/component: greptimedb-standalone
+    spec:
+      volumes:
+        - name: logs
+          emptyDir: {}
+        - name: config
+          configMap:
+            name: greptimedb-standalone
+            defaultMode: 420
+      containers:
+        - name: standalone
+          image: docker.io/greptime/greptimedb:latest
+          args:
+            - standalone
+            - start
+            - '--rpc-bind-addr'
+            - 0.0.0.0:4001
+            - '--mysql-addr'
+            - 0.0.0.0:4002
+            - '--http-addr'
+            - 0.0.0.0:4000
+            - '--postgres-addr'
+            - 0.0.0.0:4003
+            - '--config-file'
+            - /etc/greptimedb/config.toml
+          ports:
+            - name: grpc
+              containerPort: 4001
+              protocol: TCP
+            - name: http
+              containerPort: 4000
+              protocol: TCP
+            - name: mysql
+              containerPort: 4002
+              protocol: TCP
+            - name: postgres
+              containerPort: 4003
+              protocol: TCP
+          resources: {}
+          volumeMounts:
+            - name: datanode
+              mountPath: /data/greptimedb
+            - name: logs
+              mountPath: /data/greptimedb/logs
+            - name: config
+              mountPath: /etc/greptimedb
+          livenessProbe:
+            httpGet:
+              path: /health
+              port: 4000
+              scheme: HTTP
+            timeoutSeconds: 1
+            periodSeconds: 5
+            successThreshold: 1
+            failureThreshold: 10
+          readinessProbe:
+            httpGet:
+              path: /health
+              port: 4000
+              scheme: HTTP
+            timeoutSeconds: 1
+            periodSeconds: 5
+            successThreshold: 1
+            failureThreshold: 10
+          startupProbe:
+            httpGet:
+              path: /health
+              port: 4000
+              scheme: HTTP
+            timeoutSeconds: 1
+            periodSeconds: 5
+            successThreshold: 1
+            failureThreshold: 60
+          imagePullPolicy: Always
+      restartPolicy: Always
+      terminationGracePeriodSeconds: 30
+      dnsPolicy: ClusterFirst
+  volumeClaimTemplates:
+    - kind: PersistentVolumeClaim
+      apiVersion: v1
+      metadata:
+        name: datanode
+        creationTimestamp: null
+      spec:
+        accessModes:
+          - ReadWriteOnce
+        resources:
+          requests:
+            storage: 20Gi
+        volumeMode: Filesystem
+  serviceName: ''
+  podManagementPolicy: OrderedReady
+  updateStrategy:
+    type: RollingUpdate
+    rollingUpdate:
+      partition: 0
+  revisionHistoryLimit: 10
+  persistentVolumeClaimRetentionPolicy:
+    whenDeleted: Retain
+    whenScaled: Retain
 {{- end }}