mdstat for python

This short library is used to convert your /proc/mdstat file into an object usable in Python. It comes with a short mdjson script that dumps your file in a JSON format.

Requirements

• Linux (or at least get mdstat files from a Linux machine)
• Python 2.6+ (or 3.2+)

Example usage

For most use cases, you can simple run mdstat.parse() to get nested dictionaries repesenting your local /proc/mdstat file. If you fetch the file remotely or need to run the parser on a stream, use parse_stream.

$ mdjson
{
    "personalities": [
        "raid10"
    ],
    "unused_devices": [],
    "devices": {
        "md127": {
            "read_only": true,
            "status": {
                "near_copies": 2,
                "blocks": 34359475200,
                "raid_disks": 4,
                "offset_copies": 0,
                "far_copies": 1,
                "synced": [
                    true,
                    true,
                    true,
                    true
                ],
                "chunks": "512K",
                "super": "1.2",
                "non_degraded_disks": 4
            },
            "bitmap": null,
            "resync": null,
            "active": true,
            "disks": {
                "xvdf": {
                    "spare": true,
                    "faulty": false,
                    "write_mostly": false,
                    "number": 4,
                    "replacement": false
                },
                "xvdd": {
                    "spare": false,
                    "faulty": false,
                    "write_mostly": false,
                    "number": 2,
                    "replacement": false
                },
                "xvde": {
                    "spare": false,
                    "faulty": false,
                    "write_mostly": false,
                    "number": 3,
                    "replacement": false
                },
                "xvdb": {
                    "spare": false,
                    "faulty": false,
                    "write_mostly": false,
                    "number": 0,
                    "replacement": false
                },
                "xvdc": {
                    "spare": false,
                    "faulty": false,
                    "write_mostly": false,
                    "number": 1,
                    "replacement": false
                }
            },
            "personality": "raid10"
        }
    }
}

You can also use the python interface to generate a nested dictionary:

>>> import mdstat
>>> mdstat.parse()
{
    "personalities": [
        "raid1",
        "raid5",
        "raid10",
    ],
    "devices": {
        "md0": {
            "active": True,
            [... snip ...]
        }
    }
    "unused_devices": []
}