[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6C49TTAVC8Q9CTTC05F6VSR":3,"skill-same:01M6C49TTAVC8Q9CTTC05F6VSR":86},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":31,"publishedAt":51,"linkOnly":52,"markdownContent":53,"markdownBody":54,"frontMatter":55,"refName":64,"commitSha":65,"skillDirectory":5,"forksCount":66,"repositoryHtmlUrl":67,"repositoryHomepageUrl":6,"ownerHtmlUrl":68,"ownerBlogUrl":69,"ownerDescription":70,"updatedAt":71,"install":72,"resources":85},"01M6C49TTAVC8Q9CTTC05F6VSR","vllm-deployment",null,"Deploy vLLM for high-performance LLM inference. Covers Docker CPU\u002FGPU deployments and cloud VM provisioning with OpenAI-compatible API endpoints.","stakpak\u002Fcommunity-paks","stakpak","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F86326525?v=4","01M6C48005ADA10YG0BQQX0J4Q","Stakpak","01M6C4804JD6KKZR9AQ9TYS0BT",4,"MIT","https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Fvllm-deployment\u002FSKILL.md",[18,23,26],{"categoryNo":19,"categoryName":20,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000010","后端与接口","SCENARIO","使用场景",{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},"01JBSEEDCAT000000000000011","运维与部署",{"categoryNo":27,"categoryName":28,"taxonomyType":29,"taxonomyTypeDesc":30},"01JBSEEDCAT000000000000030","开发工程师","PROFESSION","适用职业",[32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50],"vllm deployment","deployment","vLLM","cloud","openai-api","llm inference","learning","team","api","provisioning","machine-learning","docker","gpu","openai","dev","vm","deploy","performance","inference","2026-11-10 10:12:30",false,"---\nname: vllm-deployment\ndescription: |\n  Deploy vLLM for high-performance LLM inference. Covers Docker CPU\u002FGPU deployments and cloud VM provisioning with OpenAI-compatible API endpoints.\nlicense: MIT\ntags:\n  - vllm\n  - llm\n  - inference\n  - gpu\n  - ai\n  - machine-learning\n  - docker\n  - openai-api\nmetadata:\n  author: Stakpak \u003Cteam@stakpak.dev>\n  version: \"1.0.3\"\n---\n\n# vLLM Model Serving and Inference\n\n## Quick Start\n\n### Docker (CPU)\n\n```bash\ndocker run --rm -p 8000:8000 \\\n  --shm-size=4g \\\n  --cap-add SYS_NICE \\\n  --security-opt seccomp=unconfined \\\n  -e VLLM_CPU_KVCACHE_SPACE=4 \\\n  \u003Cvllm-cpu-image> \\\n  --model \u003Cmodel-name> \\\n  --dtype float32\n# Access: http:\u002F\u002Flocalhost:8000\n```\n\n### Docker (GPU)\n\n```bash\ndocker run --rm -p 8000:8000 \\\n  --gpus all \\\n  --shm-size=4g \\\n  \u003Cvllm-gpu-image> \\\n  --model \u003Cmodel-name>\n# Access: http:\u002F\u002Flocalhost:8000\n```\n\n## Docker Deployment\n\n### 1. Assess Hardware Requirements\n\n| Hardware | Minimum RAM | Recommended |\n|----------|-------------|-------------|\n| CPU | 2x model size | 4x model size |\n| GPU | Model size + 2GB | Model size + 4GB VRAM |\n\n- Check model documentation for specific requirements\n- Consider quantized variants to reduce memory footprint\n- Allocate 50-100GB storage for model downloads\n\n### 2. Pull the Container Image\n\n```bash\n# CPU image (check vLLM docs for latest tag)\ndocker pull \u003Cvllm-cpu-image>\n\n# GPU image (check vLLM docs for latest tag)\ndocker pull \u003Cvllm-gpu-image>\n```\n\n**Notes:**\n- Use CPU-specific images for CPU inference\n- Use CUDA-enabled images matching your GPU architecture\n- Verify CPU instruction set compatibility (AVX512, AVX2)\n\n### 3. Configure and Run\n\n**CPU Deployment:**\n\n```bash\ndocker run --rm \\\n  --shm-size=4g \\\n  --cap-add SYS_NICE \\\n  --security-opt seccomp=unconfined \\\n  -p 8000:8000 \\\n  -e VLLM_CPU_KVCACHE_SPACE=4 \\\n  -e VLLM_CPU_OMP_THREADS_BIND=0-7 \\\n  \u003Cvllm-cpu-image> \\\n  --model \u003Cmodel-name> \\\n  --dtype float32 \\\n  --max-model-len 2048\n```\n\n**GPU Deployment:**\n\n```bash\ndocker run --rm \\\n  --gpus all \\\n  --shm-size=4g \\\n  -p 8000:8000 \\\n  \u003Cvllm-gpu-image> \\\n  --model \u003Cmodel-name> \\\n  --dtype auto \\\n  --max-model-len 4096\n```\n\n### 4. Verify Deployment\n\n```bash\n# Check health\ncurl http:\u002F\u002Flocalhost:8000\u002Fhealth\n\n# List models\ncurl http:\u002F\u002Flocalhost:8000\u002Fv1\u002Fmodels\n\n# Test inference\ncurl http:\u002F\u002Flocalhost:8000\u002Fv1\u002Fcompletions \\\n  -H \"Content-Type: application\u002Fjson\" \\\n  -d '{\"model\": \"\u003Cmodel-name>\", \"prompt\": \"Hello\", \"max_tokens\": 10}'\n```\n\n### 5. Update\n\n```bash\ndocker pull \u003Cvllm-image>\ndocker stop \u003Ccontainer-id>\n# Re-run with same parameters\n```\n\n## Cloud VM Deployment\n\n### 1. Provision Infrastructure\n\n```bash\n# Create security group with rules:\n# - TCP 22 (SSH)\n# - TCP 8000 (API)\n\n# Launch instance with:\n# - Sufficient RAM\u002FVRAM for model\n# - Docker pre-installed (or install manually)\n# - 50-100GB root volume\n# - Public IP for external access\n```\n\n### 2. Connect and Deploy\n\n```bash\nssh -i \u003Ckey-file> \u003Cuser>@\u003Cinstance-ip>\n\n# Install Docker if not present\n# Pull and run vLLM container (see Docker Deployment section)\n```\n\n### 3. Verify External Access\n\n```bash\n# From local machine\ncurl http:\u002F\u002F\u003Cinstance-ip>:8000\u002Fhealth\ncurl http:\u002F\u002F\u003Cinstance-ip>:8000\u002Fv1\u002Fmodels\n```\n\n### 4. Cleanup\n\n```bash\n# Stop container\ndocker stop \u003Ccontainer-id>\n\n# Terminate instance to stop costs\n# Delete associated resources (volumes, security groups) if temporary\n```\n\n## Configuration Reference\n\n### Environment Variables\n\n| Variable | Purpose | Example |\n|----------|---------|---------|\n| `VLLM_CPU_KVCACHE_SPACE` | KV cache size in GB (CPU) | `4` |\n| `VLLM_CPU_OMP_THREADS_BIND` | CPU core binding (CPU) | `0-7` |\n| `CUDA_VISIBLE_DEVICES` | GPU device selection | `0,1` |\n| `HF_TOKEN` | HuggingFace authentication | `hf_xxx` |\n\n### Docker Flags\n\n| Flag | Purpose |\n|------|---------|\n| `--shm-size=4g` | Shared memory for IPC |\n| `--cap-add SYS_NICE` | NUMA optimization (CPU) |\n| `--security-opt seccomp=unconfined` | Memory policy syscalls (CPU) |\n| `--gpus all` | GPU access |\n| `-p 8000:8000` | Port mapping |\n\n### vLLM Arguments\n\n| Argument | Purpose | Example |\n|----------|---------|---------|\n| `--model` | Model name\u002Fpath | `\u003Cmodel-name>` |\n| `--dtype` | Data type | `float32`, `auto`, `bfloat16` |\n| `--max-model-len` | Max context length | `2048` |\n| `--tensor-parallel-size` | Multi-GPU parallelism | `2` |\n\n## API Endpoints\n\n| Endpoint | Method | Purpose |\n|----------|--------|---------|\n| `\u002Fhealth` | GET | Health check |\n| `\u002Fv1\u002Fmodels` | GET | List available models |\n| `\u002Fv1\u002Fcompletions` | POST | Text completion |\n| `\u002Fv1\u002Fchat\u002Fcompletions` | POST | Chat completion |\n| `\u002Fmetrics` | GET | Prometheus metrics |\n\n## Production Checklist\n\n- [ ] Verify model fits in available memory\n- [ ] Configure appropriate data type for hardware\n- [ ] Set up firewall\u002Fsecurity group rules\n- [ ] Test API endpoints before production use\n- [ ] Configure monitoring (Prometheus metrics)\n- [ ] Set up health check alerts\n- [ ] Document model and configuration used\n- [ ] Plan for model updates and rollbacks\n\n## Troubleshooting\n\n| Issue | Solution |\n|-------|----------|\n| Container exits immediately | Increase RAM or use smaller model |\n| Slow inference (CPU) | Verify OMP thread binding configuration |\n| Connection refused externally | Check firewall\u002Fsecurity group rules |\n| Model download fails | Set HF_TOKEN for gated models |\n| Out of memory during inference | Reduce max_model_len or batch size |\n| Port already in use | Change host port mapping |\n| Warmup takes too long | Normal for large models (1-5 min) |\n\n## References\n\n- [vLLM Documentation](https:\u002F\u002Fdocs.vllm.ai\u002F)\n- [vLLM CPU Installation](https:\u002F\u002Fdocs.vllm.ai\u002Fen\u002Fstable\u002Fgetting_started\u002Finstallation\u002Fcpu\u002F)\n- [vLLM GPU Installation](https:\u002F\u002Fdocs.vllm.ai\u002Fen\u002Fstable\u002Fgetting_started\u002Finstallation\u002Fgpu\u002F)\n- [vLLM Supported Models](https:\u002F\u002Fdocs.vllm.ai\u002Fen\u002Fstable\u002Fmodels\u002Fsupported_models.html)\n- [OpenAI API Reference](https:\u002F\u002Fplatform.openai.com\u002Fdocs\u002Fapi-reference)\n- [HuggingFace Model Hub](https:\u002F\u002Fhuggingface.co\u002Fmodels)\n","# vLLM Model Serving and Inference\n\n## Quick Start\n\n### Docker (CPU)\n\n```bash\ndocker run --rm -p 8000:8000 \\\n  --shm-size=4g \\\n  --cap-add SYS_NICE \\\n  --security-opt seccomp=unconfined \\\n  -e VLLM_CPU_KVCACHE_SPACE=4 \\\n  \u003Cvllm-cpu-image> \\\n  --model \u003Cmodel-name> \\\n  --dtype float32\n# Access: http:\u002F\u002Flocalhost:8000\n```\n\n### Docker (GPU)\n\n```bash\ndocker run --rm -p 8000:8000 \\\n  --gpus all \\\n  --shm-size=4g \\\n  \u003Cvllm-gpu-image> \\\n  --model \u003Cmodel-name>\n# Access: http:\u002F\u002Flocalhost:8000\n```\n\n## Docker Deployment\n\n### 1. Assess Hardware Requirements\n\n| Hardware | Minimum RAM | Recommended |\n|----------|-------------|-------------|\n| CPU | 2x model size | 4x model size |\n| GPU | Model size + 2GB | Model size + 4GB VRAM |\n\n- Check model documentation for specific requirements\n- Consider quantized variants to reduce memory footprint\n- Allocate 50-100GB storage for model downloads\n\n### 2. Pull the Container Image\n\n```bash\n# CPU image (check vLLM docs for latest tag)\ndocker pull \u003Cvllm-cpu-image>\n\n# GPU image (check vLLM docs for latest tag)\ndocker pull \u003Cvllm-gpu-image>\n```\n\n**Notes:**\n- Use CPU-specific images for CPU inference\n- Use CUDA-enabled images matching your GPU architecture\n- Verify CPU instruction set compatibility (AVX512, AVX2)\n\n### 3. Configure and Run\n\n**CPU Deployment:**\n\n```bash\ndocker run --rm \\\n  --shm-size=4g \\\n  --cap-add SYS_NICE \\\n  --security-opt seccomp=unconfined \\\n  -p 8000:8000 \\\n  -e VLLM_CPU_KVCACHE_SPACE=4 \\\n  -e VLLM_CPU_OMP_THREADS_BIND=0-7 \\\n  \u003Cvllm-cpu-image> \\\n  --model \u003Cmodel-name> \\\n  --dtype float32 \\\n  --max-model-len 2048\n```\n\n**GPU Deployment:**\n\n```bash\ndocker run --rm \\\n  --gpus all \\\n  --shm-size=4g \\\n  -p 8000:8000 \\\n  \u003Cvllm-gpu-image> \\\n  --model \u003Cmodel-name> \\\n  --dtype auto \\\n  --max-model-len 4096\n```\n\n### 4. Verify Deployment\n\n```bash\n# Check health\ncurl http:\u002F\u002Flocalhost:8000\u002Fhealth\n\n# List models\ncurl http:\u002F\u002Flocalhost:8000\u002Fv1\u002Fmodels\n\n# Test inference\ncurl http:\u002F\u002Flocalhost:8000\u002Fv1\u002Fcompletions \\\n  -H \"Content-Type: application\u002Fjson\" \\\n  -d '{\"model\": \"\u003Cmodel-name>\", \"prompt\": \"Hello\", \"max_tokens\": 10}'\n```\n\n### 5. Update\n\n```bash\ndocker pull \u003Cvllm-image>\ndocker stop \u003Ccontainer-id>\n# Re-run with same parameters\n```\n\n## Cloud VM Deployment\n\n### 1. Provision Infrastructure\n\n```bash\n# Create security group with rules:\n# - TCP 22 (SSH)\n# - TCP 8000 (API)\n\n# Launch instance with:\n# - Sufficient RAM\u002FVRAM for model\n# - Docker pre-installed (or install manually)\n# - 50-100GB root volume\n# - Public IP for external access\n```\n\n### 2. Connect and Deploy\n\n```bash\nssh -i \u003Ckey-file> \u003Cuser>@\u003Cinstance-ip>\n\n# Install Docker if not present\n# Pull and run vLLM container (see Docker Deployment section)\n```\n\n### 3. Verify External Access\n\n```bash\n# From local machine\ncurl http:\u002F\u002F\u003Cinstance-ip>:8000\u002Fhealth\ncurl http:\u002F\u002F\u003Cinstance-ip>:8000\u002Fv1\u002Fmodels\n```\n\n### 4. Cleanup\n\n```bash\n# Stop container\ndocker stop \u003Ccontainer-id>\n\n# Terminate instance to stop costs\n# Delete associated resources (volumes, security groups) if temporary\n```\n\n## Configuration Reference\n\n### Environment Variables\n\n| Variable | Purpose | Example |\n|----------|---------|---------|\n| `VLLM_CPU_KVCACHE_SPACE` | KV cache size in GB (CPU) | `4` |\n| `VLLM_CPU_OMP_THREADS_BIND` | CPU core binding (CPU) | `0-7` |\n| `CUDA_VISIBLE_DEVICES` | GPU device selection | `0,1` |\n| `HF_TOKEN` | HuggingFace authentication | `hf_xxx` |\n\n### Docker Flags\n\n| Flag | Purpose |\n|------|---------|\n| `--shm-size=4g` | Shared memory for IPC |\n| `--cap-add SYS_NICE` | NUMA optimization (CPU) |\n| `--security-opt seccomp=unconfined` | Memory policy syscalls (CPU) |\n| `--gpus all` | GPU access |\n| `-p 8000:8000` | Port mapping |\n\n### vLLM Arguments\n\n| Argument | Purpose | Example |\n|----------|---------|---------|\n| `--model` | Model name\u002Fpath | `\u003Cmodel-name>` |\n| `--dtype` | Data type | `float32`, `auto`, `bfloat16` |\n| `--max-model-len` | Max context length | `2048` |\n| `--tensor-parallel-size` | Multi-GPU parallelism | `2` |\n\n## API Endpoints\n\n| Endpoint | Method | Purpose |\n|----------|--------|---------|\n| `\u002Fhealth` | GET | Health check |\n| `\u002Fv1\u002Fmodels` | GET | List available models |\n| `\u002Fv1\u002Fcompletions` | POST | Text completion |\n| `\u002Fv1\u002Fchat\u002Fcompletions` | POST | Chat completion |\n| `\u002Fmetrics` | GET | Prometheus metrics |\n\n## Production Checklist\n\n- [ ] Verify model fits in available memory\n- [ ] Configure appropriate data type for hardware\n- [ ] Set up firewall\u002Fsecurity group rules\n- [ ] Test API endpoints before production use\n- [ ] Configure monitoring (Prometheus metrics)\n- [ ] Set up health check alerts\n- [ ] Document model and configuration used\n- [ ] Plan for model updates and rollbacks\n\n## Troubleshooting\n\n| Issue | Solution |\n|-------|----------|\n| Container exits immediately | Increase RAM or use smaller model |\n| Slow inference (CPU) | Verify OMP thread binding configuration |\n| Connection refused externally | Check firewall\u002Fsecurity group rules |\n| Model download fails | Set HF_TOKEN for gated models |\n| Out of memory during inference | Reduce max_model_len or batch size |\n| Port already in use | Change host port mapping |\n| Warmup takes too long | Normal for large models (1-5 min) |\n\n## References\n\n- [vLLM Documentation](https:\u002F\u002Fdocs.vllm.ai\u002F)\n- [vLLM CPU Installation](https:\u002F\u002Fdocs.vllm.ai\u002Fen\u002Fstable\u002Fgetting_started\u002Finstallation\u002Fcpu\u002F)\n- [vLLM GPU Installation](https:\u002F\u002Fdocs.vllm.ai\u002Fen\u002Fstable\u002Fgetting_started\u002Finstallation\u002Fgpu\u002F)\n- [vLLM Supported Models](https:\u002F\u002Fdocs.vllm.ai\u002Fen\u002Fstable\u002Fmodels\u002Fsupported_models.html)\n- [OpenAI API Reference](https:\u002F\u002Fplatform.openai.com\u002Fdocs\u002Fapi-reference)\n- [HuggingFace Model Hub](https:\u002F\u002Fhuggingface.co\u002Fmodels)\n",{"name":5,"tags":56,"license":15,"metadata":60,"description":63},[57,58,50,44,59,42,43,36],"vllm","llm","ai",{"author":61,"version":62},"Stakpak \u003Cteam@stakpak.dev>","1.0.3","Deploy vLLM for high-performance LLM inference. Covers Docker CPU\u002FGPU deployments and cloud VM provisioning with OpenAI-compatible API endpoints.\n","main","d9bbde9885439c55adad930456b2660b0c37c052",1,"https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks","https:\u002F\u002Fgithub.com\u002Fstakpak","https:\u002F\u002Fwww.stakpak.dev","AI that runs your Infra","2026-07-04 20:18:32",{"source":8,"skillName":5,"commands":73,"prompt":83,"zipAvailable":84},[74,77,80],{"tool":75,"command":76},"npx","npx skills add stakpak\u002Fcommunity-paks --skill vllm-deployment",{"tool":78,"command":79},"bunx","bunx skills add stakpak\u002Fcommunity-paks --skill vllm-deployment",{"tool":81,"command":82},"pnpm","pnpm dlx skills add stakpak\u002Fcommunity-paks --skill vllm-deployment","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Fvllm-deployment\u002FSKILL.md",true,[],[87,111,129,149,167,178],{"skillNo":88,"name":89,"nameZh":6,"description":90,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":91,"categories":92,"keywords":94,"publishedAt":51},"01M6C49R31XN7KJ7Y0Y78XVYTQ","terrateam-usage-guide","A comprehensive guide for configuring and using Terrateam, an open-source GitOps CI\u002FCD platform for Terraform and OpenTofu. Covers configuration file structure, access control, workflows, OIDC authentication, drift detection, and best practices for managing infrastructure through pull requests.","https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Fterrateam-usage-guide\u002FSKILL.md",[93],{"categoryNo":24,"categoryName":25,"taxonomyType":21,"taxonomyTypeDesc":22},[95,96,97,98,99,100,39,101,102,103,104,105,106,46,107,108,109,110],"terrateam usage guide","guide","pull requests","ci","opentofu","access","open","detection","oidc","workflows","file","pull","terraform","best practices","authentication","infrastructure",{"skillNo":112,"name":113,"nameZh":6,"description":114,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":115,"categories":116,"keywords":117,"publishedAt":51},"01M6C49N78AMM570EGPDXEZZDY","simple-deployment-on-vm","How to do simple but secure deployments using virtual machines on different cloud providers","https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Fsimple-deployment-on-vm\u002FSKILL.md",[],[118,47,33,119,120,121,122,39,123,124,125,126,35,46,127,128,48],"simple deployment on vm","dns","tls configuration","virtual","how","ec2","Providers","tls","aws","secure","security",{"skillNo":130,"name":131,"nameZh":6,"description":132,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":133,"categories":134,"keywords":135,"publishedAt":51},"01M6C49J8PRVT623JWRTTKFNJV","nextjs-on-aws","This document compares various AWS deployment options for Next.js applications, analyzing costs, benefits, limitations, and providing recommendations based on different use cases.","https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Fnextjs-on-aws\u002FSKILL.md",[],[136,126,137,138,33,46,139,140,141,142,39,143,144,145,146,147,148],"nextjs on aws","nextjs","architecture","analyzing","applications","analysis","Next.js","js","aws deployment","use","cost","cost-analysis","migration",{"skillNo":150,"name":151,"nameZh":6,"description":152,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":153,"categories":154,"keywords":155,"publishedAt":51},"01M6C49F5B4BF2CHNT4X08XP69","migrating-bitnami-to-bitnami-legacy","This rule book helps you migrate Bitnami Helm charts and container images from the bitnami repository to the bitnamilegacy repository.\n\nThis migration is necessary due to Bitnami's transition, effective August 28th, 2025, where existing images will be moved to the legacy repo","https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Fmigrating-bitnami-to-bitnami-legacy\u002FSKILL.md",[],[156,157,158,159,160,161,162,163,39,164,165,46,166],"migrating bitnami to bitnami legacy","helm","bitnami-helm-charts","bitnami-legacy","bitnami","container","repo","migrate","Charts","repository","container images",{"skillNo":168,"name":169,"nameZh":6,"description":170,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":171,"categories":172,"keywords":173,"publishedAt":51},"01M6C49CHCPGZZ42P5W1MTV3WS","k3s-backup","Backup and recovery solution for K3s clusters. Includes scheduled, pre-shutdown, and manual backup strategies with recovery procedures.","https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Fk3s-backup\u002FSKILL.md",[],[174,175,176,46,39,177],"k3s backup","k3s","backup","recovery",{"skillNo":179,"name":180,"nameZh":6,"description":181,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":182,"categories":183,"keywords":184,"publishedAt":51},"01M6C49ANM9YZQRP7FQPEKCPBS","infrastructure-software-upgrades","Generic guidelines on how to perform infrastructure component upgrades in a reliable way","https:\u002F\u002Fgithub.com\u002Fstakpak\u002Fcommunity-paks\u002Fblob\u002Fmain\u002Finfrastructure-software-upgrades\u002FSKILL.md",[],[110,185,186,46,39,122,187,188,189,190,191],"upgrades","infrastructure software upgrades","component","guidelines","maintenance","devops","operations"]