{
  "id": "cuda/mps-max-partition-size-exceeded",
  "signature": "CUDA error: MPS server: maximum partition size exceeded (cudaErrorMpsMaxPartitionSizeExceeded)",
  "signature_zh": "CUDA 错误：MPS 服务器：超出最大分区大小 (cudaErrorMpsMaxPartitionSizeExceeded)",
  "regex": "CUDA error: MPS server: maximum partition size exceeded",
  "domain": "cuda",
  "category": "resource_error",
  "subcategory": null,
  "root_cause": "The CUDA Multi-Process Service (MPS) server has reached its configured maximum partition size, preventing new client connections or memory allocations.",
  "root_cause_type": "generic",
  "root_cause_zh": "CUDA 多进程服务 (MPS) 服务器已达到其配置的最大分区大小，阻止了新客户端连接或内存分配。",
  "versions": [
    {
      "version": "CUDA 11.8",
      "introduced": null,
      "deprecated": null,
      "removed": null,
      "behavior_change": null,
      "status": "active"
    },
    {
      "version": "CUDA 12.1",
      "introduced": null,
      "deprecated": null,
      "removed": null,
      "behavior_change": null,
      "status": "active"
    },
    {
      "version": "CUDA 12.3",
      "introduced": null,
      "deprecated": null,
      "removed": null,
      "behavior_change": null,
      "status": "active"
    }
  ],
  "os_specific": {},
  "dead_ends": [
    {
      "action": "",
      "why_fails": "Rebooting the node resets MPS but loses all running jobs and doesn't fix the underlying configuration issue.",
      "fail_rate": 0.7,
      "condition": "",
      "sources": []
    },
    {
      "action": "",
      "why_fails": "Setting CUDA_MPS_PIPE_DIRECTORY to a temp path without restarting the MPS daemon has no effect on the partition size limit.",
      "fail_rate": 0.9,
      "condition": "",
      "sources": []
    }
  ],
  "workarounds": [
    {
      "action": "Restart the MPS daemon with a larger partition size (e.g., 40GB) using nvidia-cuda-mps-control. Log in as root and run: echo 'set_default_active_thread_percentage 100' | nvidia-cuda-mps-control; echo 'set_default_partition_size 40000MB' | nvidia-cuda-mps-control; then restart client processes.",
      "success_rate": 0.8,
      "how": "Restart the MPS daemon with a larger partition size (e.g., 40GB) using nvidia-cuda-mps-control. Log in as root and run: echo 'set_default_active_thread_percentage 100' | nvidia-cuda-mps-control; echo 'set_default_partition_size 40000MB' | nvidia-cuda-mps-control; then restart client processes.",
      "condition": "",
      "sources": []
    },
    {
      "action": "Increase the partition size via environment variable before starting the MPS server: export CUDA_MPS_PARTITION_SIZE=40000 (in MB), then restart the MPS daemon with 'nvidia-cuda-mps-control -d'.",
      "success_rate": 0.75,
      "how": "Increase the partition size via environment variable before starting the MPS server: export CUDA_MPS_PARTITION_SIZE=40000 (in MB), then restart the MPS daemon with 'nvidia-cuda-mps-control -d'.",
      "condition": "",
      "sources": []
    }
  ],
  "workarounds_zh": [
    "Restart the MPS daemon with a larger partition size (e.g., 40GB) using nvidia-cuda-mps-control. Log in as root and run: echo 'set_default_active_thread_percentage 100' | nvidia-cuda-mps-control; echo 'set_default_partition_size 40000MB' | nvidia-cuda-mps-control; then restart client processes.",
    "Increase the partition size via environment variable before starting the MPS server: export CUDA_MPS_PARTITION_SIZE=40000 (in MB), then restart the MPS daemon with 'nvidia-cuda-mps-control -d'."
  ],
  "transition_graph": {
    "leads_to": [],
    "preceded_by": [],
    "frequently_confused_with": []
  },
  "official_doc_url": "https://docs.nvidia.com/deploy/mps/index.html#topic_5_3",
  "official_doc_section": null,
  "error_code": "cudaErrorMpsMaxPartitionSizeExceeded",
  "verification_tier": "ai_generated",
  "confidence": 0.85,
  "fix_success_rate": 0.75,
  "resolvable": "true",
  "first_seen": "2024-03-15",
  "last_confirmed": "2024-06-01",
  "last_updated": "2024-06-01",
  "evidence_count": 1,
  "tags": [],
  "locale": "en",
  "aliases": []
}