Skip to content

server_info

resource_tracker.server_info #

Detect server hardware (CPU count, memory amount, disk space, GPU count and VRAM amount) via procfs or psutil, and nvidia-smi.

Functions:

Name Description
get_total_memory_mb

Get total system memory in MB from /proc/meminfo or using psutil.
get_gpu_info

Get GPU information using nvidia-smi command.
get_server_info

Collects important information about the Linux server.

get_total_memory_mb #

get_total_memory_mb()

Get total system memory in MB from /proc/meminfo or using psutil.

Source code in resource_tracker/server_info.py 
11
12
13
14
15
16
17
18
19
20
21
22
23
24
def get_total_memory_mb() -> float:
    """Get total system memory in MB from `/proc/meminfo` or using `psutil`."""
    with suppress(Exception):
        with open("/proc/meminfo", "r") as f:
            for line in f:
                if "MemTotal" in line:
                    parts = line.split(":")
                    kb = int(parts[1].strip().split()[0])
                    return round(kb / (1024), 2)
    with suppress(Exception):
        from psutil import virtual_memory

        return round(virtual_memory().total / (1024**2), 2)
    return 0

get_gpu_info #

get_gpu_info()

Get GPU information using nvidia-smi command.

Returns:

Type Description
dict

A dictionary containing GPU information:

  • count: Number of GPUs
  • memory_mb: Total VRAM in MB
  • gpu_names: List of GPU names
Source code in resource_tracker/server_info.py 
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
def get_gpu_info() -> dict:
    """Get GPU information using `nvidia-smi` command.

    Returns:
        A dictionary containing GPU information:

            - `count`: Number of GPUs
            - `memory_mb`: Total VRAM in MB
            - `gpu_names`: List of GPU names
    """
    result = {"count": 0, "memory_mb": 0, "gpu_names": []}

    with suppress(Exception):
        nvidia_smi_output = check_output(
            [
                "nvidia-smi",
                "--query-gpu=gpu_name,memory.total",
                "--format=csv,noheader,nounits",
            ],
            universal_newlines=True,
        )

        lines = nvidia_smi_output.strip().split("\n")
        result["count"] = len(lines)

        total_memory_mb = 0
        for line in lines:
            if line.strip():
                parts = line.split(",")
                memory_mb = float(parts[1].strip())
                total_memory_mb += memory_mb
                result["gpu_names"].append(parts[0].strip())

        result["memory_mb"] = total_memory_mb

    return result

get_server_info #

get_server_info()

Collects important information about the Linux server.

Returns:

Type Description
dict

A dictionary containing server information:

  • os: Operating system
  • vcpus: Number of virtual CPUs
  • memory_mb: Total memory in MB
  • gpu_count: Number of GPUs (0 if not available)
  • gpu_memory_mb: Total VRAM in MB (0 if not available)
Source code in resource_tracker/server_info.py 
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
def get_server_info() -> dict:
    """
    Collects important information about the Linux server.

    Returns:
        A dictionary containing server information:

            - `os`: Operating system
            - `vcpus`: Number of virtual CPUs
            - `memory_mb`: Total memory in MB
            - `gpu_count`: Number of GPUs (`0` if not available)
            - `gpu_memory_mb`: Total VRAM in MB (`0` if not available)
    """
    gpu_info = get_gpu_info()
    info = {
        "os": system(),
        "vcpus": cpu_count(),
        "memory_mb": get_total_memory_mb(),
        "gpu_count": gpu_info["count"],
        "gpu_names": gpu_info["gpu_names"],
        "gpu_memory_mb": gpu_info["memory_mb"],
    }
    return info