Script to manually clean expired objects from GlusterFS volume.

clean_expired.py

#!/usr/bin/env python

import os
import argparse
import math
import time
import xattr
import cPickle

METADATA_KEY = 'user.swift.metadata'


def read_metadata(path):
    metadata = ''
    key = 0
    try:
        while True:
            metadata += xattr.getxattr(path, '%s%s' %
                                       (METADATA_KEY, (key or '')))
            key += 1
    except IOError:
        pass
    return cPickle.loads(metadata)


def delete_file(path):
    try:
        os.unlink(path)
    except OSError as e:
        if e.errno != errno.ENOENT:
            print "delete_file(%s): %s" % (path, e.strerror)
        


USAGE = "%(prog)s [options] <mount>"
DESCRIPTION="Delete expired objects by crawling the filsystem. \
            Expired objects are only printed, unless -d option is specified \
            which actually deletes objects from filesystem."

parser = argparse.ArgumentParser(usage=USAGE, description=DESCRIPTION)
parser.add_argument("-d", "--delete", dest="delete_objects",
                    action="store_true", default=False,
                    help="Delete expired objects found in filesystem.")
parser.add_argument("-p", "--print", dest="print_objects",
                    action="store_true", default=True,
                    help="Print expired objects found in filesystem (Default).")
parser.add_argument("mount", help="Mount point path of a GlusterFS volume.")
args = parser.parse_args()

if not os.path.isdir(args.mount):
    parser.error("Invalid mount point path %s" % args.mount)

count = 0

for root, dirs, files in os.walk(args.mount):
    for f in files:
        path = os.path.join(root, f)
        md = read_metadata(path)
        if md:
            x_delete_at = md.get('X-Delete-At', None)
            if x_delete_at:
                if int(math.floor(time.time())) > int(x_delete_at):
                    count += 1
                    if args.print_objects:
                        print path
                    if args.delete_objects:
                        delete_file(path)

print "Total number of expired objects found: %d" % (count)