diff options
Diffstat (limited to 'tools/glusterfind/src/utils.py')
| -rw-r--r-- | tools/glusterfind/src/utils.py | 180 |
1 files changed, 122 insertions, 58 deletions
diff --git a/tools/glusterfind/src/utils.py b/tools/glusterfind/src/utils.py index de9c027e299..906ebd8f252 100644 --- a/tools/glusterfind/src/utils.py +++ b/tools/glusterfind/src/utils.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python +# -*- coding: utf-8 -*- # Copyright (c) 2015 Red Hat, Inc. <http://www.redhat.com/> # This file is part of GlusterFS. @@ -9,14 +9,45 @@ # cases as published by the Free Software Foundation. import sys -import socket from subprocess import PIPE, Popen -from errno import EPERM, EEXIST +from errno import EEXIST, ENOENT +import xml.etree.cElementTree as etree import logging import os from datetime import datetime ROOT_GFID = "00000000-0000-0000-0000-000000000001" +DEFAULT_CHANGELOG_INTERVAL = 15 +SPACE_ESCAPE_CHAR = "%20" +NEWLINE_ESCAPE_CHAR = "%0A" +PERCENTAGE_ESCAPE_CHAR = "%25" + +ParseError = etree.ParseError if hasattr(etree, 'ParseError') else SyntaxError +cache_data = {} + + +class RecordType(object): + NEW = "NEW" + MODIFY = "MODIFY" + RENAME = "RENAME" + DELETE = "DELETE" + + +def cache_output(func): + def wrapper(*args, **kwargs): + global cache_data + if cache_data.get(func.__name__, None) is None: + cache_data[func.__name__] = func(*args, **kwargs) + + return cache_data[func.__name__] + return wrapper + + +def handle_rm_error(func, path, exc_info): + if exc_info[1].errno == ENOENT: + return + + raise exc_info[1] def find(path, callback_func=lambda x: True, filter_func=lambda x: True, @@ -24,30 +55,43 @@ def find(path, callback_func=lambda x: True, filter_func=lambda x: True, if path in ignore_dirs: return - if filter_func(path): - callback_func(path) + # Capture filter_func output and pass it to callback function + filter_result = filter_func(path) + if filter_result is not None: + callback_func(path, filter_result, os.path.isdir(path)) for p in os.listdir(path): full_path = os.path.join(path, p) - if os.path.isdir(full_path): + is_dir = os.path.isdir(full_path) + if is_dir: if subdirs_crawl: find(full_path, callback_func, filter_func, ignore_dirs) else: - if filter_func(full_path): - callback_func(full_path) + filter_result = filter_func(full_path) + if filter_result is not None: + callback_func(full_path, filter_result) else: - if filter_func(full_path): - callback_func(full_path) + filter_result = filter_func(full_path) + if filter_result is not None: + callback_func(full_path, filter_result, is_dir) -def output_write(f, path, prefix="."): +def output_write(f, path, prefix=".", encode=False, tag="", + field_separator=" "): if path == "": return if prefix != ".": path = os.path.join(prefix, path) - f.write("%s\n" % path) + + if encode: + path = quote_plus_space_newline(path) + + # set the field separator + FS = "" if tag == "" else field_separator + + f.write("%s%s%s\n" % (tag.strip(), FS, path)) def human_time(ts): @@ -153,51 +197,71 @@ def symlink_gfid_to_path(brick, gfid): return out_path -def is_host_local(host): +@cache_output +def get_my_uuid(): + cmd = ["gluster", "system::", "uuid", "get", "--xml"] + rc, out, err = execute(cmd) + + if rc != 0: + return None + + tree = etree.fromstring(out) + uuid_el = tree.find("uuidGenerate/uuid") + return uuid_el.text + + +def is_host_local(host_uuid): + # Get UUID only if it is not done previously + # else Cache the UUID value + my_uuid = get_my_uuid() + if my_uuid == host_uuid: + return True + + return False + + +def get_changelog_rollover_time(volumename): + cmd = ["gluster", "volume", "get", volumename, + "changelog.rollover-time", "--xml"] + rc, out, err = execute(cmd) + + if rc != 0: + return DEFAULT_CHANGELOG_INTERVAL + + try: + tree = etree.fromstring(out) + val = tree.find('volGetopts/Opt/Value').text + if val is not None: + # Filter the value by split, as it may be 'X (DEFAULT)' + # and we only need 'X' + return int(val.split(' ', 1)[0]) + except ParseError: + return DEFAULT_CHANGELOG_INTERVAL + + +def output_path_prepare(path, args): """ - Find if a host is local or not. - Code copied from $GLUSTERFS/geo-replication/syncdaemon/syncdutils.py + If Prefix is set, joins to Path, removes ending slash + and encodes it. """ - locaddr = False - for ai in socket.getaddrinfo(host, None): - # cf. http://github.com/gluster/glusterfs/blob/ce111f47/xlators - # /mgmt/glusterd/src/glusterd-utils.c#L125 - if ai[0] == socket.AF_INET: - if ai[-1][0].split(".")[0] == "127": - locaddr = True - break - elif ai[0] == socket.AF_INET6: - if ai[-1][0] == "::1": - locaddr = True - break - else: - continue - try: - # use ICMP socket to avoid net.ipv4.ip_nonlocal_bind issue, - # cf. https://bugzilla.redhat.com/show_bug.cgi?id=890587 - s = socket.socket(ai[0], socket.SOCK_RAW, socket.IPPROTO_ICMP) - except socket.error: - ex = sys.exc_info()[1] - if ex.errno != EPERM: - raise - f = None - try: - f = open("/proc/sys/net/ipv4/ip_nonlocal_bind") - if int(f.read()) != 0: - logger.warning("non-local bind is set and not " - "allowed to create " - "raw sockets, cannot determine " - "if %s is local" % host) - return False - s = socket.socket(ai[0], socket.SOCK_DGRAM) - finally: - if f: - f.close() - try: - s.bind(ai[-1]) - locaddr = True - break - except: - pass - s.close() - return locaddr + if args.output_prefix != ".": + path = os.path.join(args.output_prefix, path) + if path.endswith("/"): + path = path[0:len(path)-1] + + if args.no_encode: + return path + else: + return quote_plus_space_newline(path) + + +def unquote_plus_space_newline(s): + return s.replace(SPACE_ESCAPE_CHAR, " ")\ + .replace(NEWLINE_ESCAPE_CHAR, "\n")\ + .replace(PERCENTAGE_ESCAPE_CHAR, "%") + + +def quote_plus_space_newline(s): + return s.replace("%", PERCENTAGE_ESCAPE_CHAR)\ + .replace(" ", SPACE_ESCAPE_CHAR)\ + .replace("\n", NEWLINE_ESCAPE_CHAR) |