1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
|
# Copyright (C) 2020 Xin Liang <XLiang@suse.com>
# See COPYING for license information.
import re
import time
from contextlib import contextmanager
from . import sh
from . import config
from . import log
from .sh import ShellUtils
logger = log.setup_logger(__name__)
class SSHError(Exception):
"""
Custom exception for ssh error
"""
class ClaimLockError(Exception):
"""
Custom exception if claiming lock failed or wait lock release timed out
"""
class Lock(object):
"""
A base class define a lock mechanism used to exclude other nodes
"""
LOCK_DIR_DEFAULT = "/run/.crmsh_lock_directory"
def __init__(self, lock_dir=None):
"""
Init function
"""
# only the lock owner can unlock
self.lock_owner = False
self.lock_dir = lock_dir or self.LOCK_DIR_DEFAULT
def _run(self, cmd):
"""
Run command on local
"""
return ShellUtils().get_stdout_stderr(cmd)
def _create_lock_dir(self):
"""
Create lock directory, mkdir command was atomic
"""
cmd = "mkdir {}".format(self.lock_dir)
rc, _, _ = self._run(cmd)
if rc == 0:
self.lock_owner = True
return True
return False
def _lock_or_fail(self):
"""
Raise ClaimLockError if claiming lock failed
"""
if not self._create_lock_dir():
raise ClaimLockError("Failed to claim lock (the lock directory exists at {})".format(self.lock_dir))
def _unlock(self):
"""
Remove the lock directory
"""
if self.lock_owner:
cmd = "rm -rf {}".format(self.lock_dir)
self._run(cmd)
@contextmanager
def lock(self):
"""
Create lock directory on local, and remove it finally
Might raise ClaimLockError
"""
try:
self._lock_or_fail()
yield
except:
raise
finally:
self._unlock()
class RemoteLock(Lock):
"""
A class inherited from Lock class
Define the behavior how to claim lock on remote node and how to wait the lock released
"""
SSH_TIMEOUT = 10
SSH_OPTION = "-o ConnectTimeout={} -o StrictHostKeyChecking=no".format(SSH_TIMEOUT)
SSH_EXIT_ERR = 255
MIN_LOCK_TIMEOUT = 120
WAIT_INTERVAL = 10
def __init__(self, remote_node, for_join=True, lock_dir=None, wait=True, no_warn=False):
"""
Init function
"""
self.remote_node = remote_node
self.for_join = for_join
self.wait = wait
self.no_warn = no_warn
super(__class__, self).__init__(lock_dir=lock_dir)
def _run(self, cmd):
"""
Run command on remote node
"""
# TODO: pass SSH_OPTION
rc, out, err = sh.cluster_shell().get_rc_stdout_stderr_without_input(self.remote_node, cmd)
if rc == self.SSH_EXIT_ERR:
raise SSHError(err)
return rc, out, err
@property
def lock_timeout(self):
"""
Get lock_timeout from config.core
"""
try:
value = int(config.core.lock_timeout)
except ValueError:
raise ValueError("Invalid format of core.lock_timeout(should be a number)")
if value < self.MIN_LOCK_TIMEOUT:
raise ValueError("Minimum value of core.lock_timeout should be {}".format(self.MIN_LOCK_TIMEOUT))
return value
def _get_online_nodelist(self):
"""
Get the online node list from remote node
"""
rc, out, err = self._run("crm_node -l")
if rc != 0 and err:
raise ValueError(err)
return re.findall('[0-9]+ (.*) member', out)
def _lock_or_wait(self):
"""
Try to claim lock on remote node, wait if failed to claim
raise ClaimLockError if reached the lock_timeout
"""
warned_once = False
online_list = []
pre_online_list = []
expired_error_str = "Cannot continue since the lock directory exists at the node ({}:{})".format(self.remote_node, self.lock_dir)
current_time = int(time.time())
timeout = current_time + self.lock_timeout
while current_time <= timeout:
# Try to claim the lock
if self._create_lock_dir():
# Success
break
if self.for_join:
# Might lose claiming lock again, start to wait again
online_list = self._get_online_nodelist()
if pre_online_list and pre_online_list != online_list:
timeout = current_time + self.lock_timeout
pre_online_list = online_list
if not self.no_warn and not warned_once:
warned_once = True
logger.warning("Might have unfinished process on other nodes, wait %ss...", self.lock_timeout)
time.sleep(self.WAIT_INTERVAL)
current_time = int(time.time())
else:
raise ClaimLockError("Timed out after {} seconds. {}".format(self.lock_timeout, expired_error_str))
@contextmanager
def lock(self):
"""
Create lock directory on remote, and remove it finally
Might raise SSHError, ClaimLockError and ValueError
"""
try:
if self.wait:
self._lock_or_wait()
else:
self._lock_or_fail()
yield
except:
raise
finally:
self._unlock()
|
