SSTable Tombstone Counter

tc.py

import fileinput, re, operator
from collections import Counter

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():
    keys = Counter()
    data = 0
    top = 25

    for line in fileinput.input():
        current = None
        if '{"key"' in line:
            current = re.findall(': "(.*)",', line)[0]
        if '"t"' in line:
            keys[current] += 1

        data = data + len(line)

    sorted_tombstones = sorted(keys.items(), key=operator.itemgetter(1))
    sorted_tombstones.reverse()

    print 'Read {0} keys and {1} of data'.format(len(sorted_tombstones), sizeof_fmt(data))
    print 'Top {0} keys with highest number tombstones'.format(top)
    n = 0
    for pair in sorted_tombstones[0:top]:
        n += 1
        print "{0:3} {1} => {2}".format(str(n) + '.', pair[0], pair[1])

if __name__ == '__main__':
    main()

usage:

sstable2json /path/to/sstable-Data.db | python tc.py