Remove old cassandra and skylla snapshots

clean_skylla_snapshots.py

import os
import re
from datetime import datetime
from subprocess import run, PIPE


TARGET_KEYSPACE = 'prod'

_10GB = 10 ** 10

listsnapshots_re = re.compile(
    r'(?P<timestamp>\d+)\s+'
    r'(?P<keyspace>\w+)\s+'
    r'(?P<column_family_name>\S+)\s+'
    r'(?P<true_size>\d+.?\d*\s\S+)\s+'
    r'(?P<disk_size>\d+.?\d*\s\S+)'
)


units = {
    'bytes': 1,
    'KB': 10 ** 3,
    'MB': 10 ** 6,
    'GB': 10 ** 9,
    'TB': 10 ** 12,
}


def parse_size(size: str):
    number, unit = [string.strip() for string in size.split()]
    return int(float(number)*units[unit])


def sizeof_fmt(num, suffix='B'):
    for unit in ['','Ki','Mi','Gi','Ti','Pi','Ei','Zi']:
        if abs(num) < 1024.0:
            return "%3.1f%s%s" % (num, unit, suffix)
        num /= 1024.0
    return "%.1f%s%s" % (num, 'Yi', suffix)


def main():
    proc = run(['nodetool', 'listsnapshots'], stdout=PIPE, stderr=PIPE)
    out = proc.stdout.decode().split(os.linesep)
    snapshots = []
    for line in out:
        re_match = listsnapshots_re.match(line)
        if not re_match:
            continue

        group = re_match.groupdict()

        if group['keyspace'] != TARGET_KEYSPACE:
            continue

        when = datetime.fromtimestamp(float(group['timestamp']) / 1000)
        true_size = parse_size(group['true_size'])
        disk_size = parse_size(group['disk_size'])

        snapshots.append({
            'when': when,
            'timestamp': group['timestamp'],
            'true_size': true_size,
            'disk_size': disk_size,
        })
        
    true_size = sum(ss['true_size'] for ss in snapshots)
    disk_size = sum(ss['disk_size'] for ss in snapshots)
    
    disk_status = os.statvfs('/var')

    print('SNAPHOTS: {snapshots}, TRUE_SIZE: {true_size}, DISK_SIZE: {disk_size}, SPACE_AVAILABLE: {disk_available}'.format(
        snapshots=len(snapshots),
        true_size=sizeof_fmt(true_size),
        disk_size=sizeof_fmt(disk_size),
        disk_available=sizeof_fmt(disk_status.f_frsize * disk_status.f_bavail),
    ))
    
    while True:
        disk_status = os.statvfs('/var')
        if disk_status.f_frsize * disk_status.f_bavail > _10GB:
            return
        if not snapshots:
            return
        latest_snapshot = snapshots.pop()
        run(['nodetool', 'clearsnapshot', '-t', latest_snapshot['timestamp']], stdout=PIPE, stderr=PIPE)
        print('SNAPSOT DELETED', latest_snapshot)


if __name__ == '__main__':
    main()