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
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
|
# Copyright (c) 2009-2010 Gluster, Inc. <http://www.gluster.com>
# This file is part of GlusterFS.
# GlusterFS is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published
# by the Free Software Foundation; either version 3 of the License,
# or (at your option) any later version.
# GlusterFS is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
# General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with this program. If not, see
# <http://www.gnu.org/licenses/>.
import os, sys, string
#Cachesize calculator
cache_size = "`echo $(( $(grep 'MemTotal' /proc/meminfo | sed 's/[^0-9]//g') / 5120 ))`"
class CreateVolfile:
def __init__ (self, server_dict, server, transport,
transports, options, server_array):
self.host_dict = server_dict
self.host = server
self.volume_name = options.volume_name
self.transport = transport
self.transports = transports
self.auth_parameters = options.auth_param
self.raid_type = options.raid_type
self.ib_devport = options.ib_dev
self.conf_dir = options.conf_dir
self.host_array = server_array
self.enable_safe_mode = options.enable_safe_mode
self.volume_size_server = options.size_server
self.volume_size_client = options.size_client
self.nfs = options.need_nfs
self.num_replica = options.num_replica
self.num_stripe = options.num_stripe
if options.port:
self.gfs_port = options.port
self.gfs_ib_port = options.port + 1
else:
self.gfs_port = None
self.gfs_ib_port = None
def create_mount_volfile (self):
raid_type = self.raid_type
mount_volfile = "%s/%s-%s.vol" % (self.conf_dir, str(self.volume_name), str(self.transport))
mount_fd = file ("%s" % (mount_volfile), "w")
print "Generating client volfiles.. '%s'" % mount_volfile
cmdline = string.join (sys.argv, ' ')
mount_fd.write ("## file auto generated by %s\n" % sys.argv[0])
mount_fd.write ("# Cmd line:\n")
mount_fd.write ("# $ %s\n\n" % cmdline)
if raid_type is not None:
# Used for later usage
mount_fd.write ("# RAID %d\n" % raid_type)
mount_fd.write ("# TRANSPORT-TYPE %s\n" % self.transport)
subvolumes = []
for host in self.host_dict.keys():
i = 1
for export in self.host_dict[host]:
mount_fd.write ("volume %s-%s\n" % (host,i))
mount_fd.write (" type protocol/client\n")
mount_fd.write (" option transport-type %s\n" %
self.transport)
mount_fd.write (" option remote-host %s\n" % host)
if self.transport == 'ib-verbs':
mount_fd.write (" option transport.ib-verbs.port %d\n" %
self.ib_devport)
if self.gfs_ib_port:
mount_fd.write (" option remote-port %d\n" %
self.gfs_ib_port)
if self.transport == 'tcp':
mount_fd.write (" option transport.socket.nodelay on\n")
if self.gfs_port:
mount_fd.write (" option remote-port %d\n" % self.gfs_port)
mount_fd.write (" option remote-subvolume %s\n" % export)
mount_fd.write ("end-volume\n\n")
i += 1
exportlist = {}
for entry in self.host_array:
node = entry.split(':')[0]
if not exportlist.has_key(node):
exportlist[node] = 1
else:
exportlist[node] += 1
subvolumes.append(str(node) + '-' + str(exportlist[node]))
# Stripe section.. if given
if raid_type is 0:
max_stripe_idx = len (subvolumes) / self.num_stripe
stripe_idx = 0
index = 0
while index < max_stripe_idx:
mount_fd.write ("volume stripe-%d\n" % index)
mount_fd.write (" type cluster/stripe\n")
mount_fd.write ("# option block-size 128k\n")
mount_fd.write ("# option use-xattr no\n")
mount_fd.write (" subvolumes %s" % subvolumes[stripe_idx])
sub_idx = 1
while sub_idx < self.num_stripe:
mount_fd.write (" %s" % subvolumes[stripe_idx+sub_idx])
sub_idx += 1
mount_fd.write ("\nend-volume\n\n")
stripe_idx += self.num_stripe
index +=1
# Replicate section
if raid_type is 1:
max_mirror_idx = len (subvolumes) / self.num_replica
mirror_idx = 0
index = 0
while index < max_mirror_idx:
mount_fd.write ("volume mirror-%d\n" % index)
mount_fd.write (" type cluster/replicate\n")
mount_fd.write (" subvolumes %s" % subvolumes[mirror_idx])
sub_idx = 1
while sub_idx < self.num_replica:
mount_fd.write (" %s" % subvolumes[mirror_idx + sub_idx])
sub_idx += 1
mount_fd.write ("\nend-volume\n\n")
mirror_idx += self.num_replica
index += 1
# Distribute section
if raid_type is 0:
subvolumes = []
flag = 0
while flag < index:
subvolumes.append ("stripe-%d" % flag)
flag += 1
if raid_type is 1:
subvolumes = []
flag = 0
while flag < index:
subvolumes.append ("mirror-%d" % flag)
flag += 1
if len (subvolumes) > 1:
mount_fd.write ("volume distribute\n")
mount_fd.write (" type cluster/distribute\n")
mount_fd.write ("# option unhashed-sticky-bit yes"
" # Used for migrating data while adding new nodes\n")
mount_fd.write ("# option min-free-disk 5%"
" # Minimum free disk available on the volume\n")
mount_fd.write (" subvolumes %s\n" %
string.join (subvolumes,' '))
mount_fd.write ("end-volume\n\n")
subvolumes[0] = "distribute"
if self.nfs:
mount_fd.write ("volume nfsxlator\n")
mount_fd.write (" type nfs/server\n")
mount_fd.write (" subvolumes %s\n" % subvolumes[0])
mount_fd.write ("# option rpc-auth.auth-unix off #Enabled by default\n")
mount_fd.write ("# option rpc-auth.auth-null off #Enabled by default\n")
mount_fd.write ("# By default all addresses are rejected until allowed.\n")
mount_fd.write ("# option rpc-auth.addr.reject 127.*\n")
mount_fd.write ("# option rpc-auth.addr.allow localhost\n")
mount_fd.write (" option rpc-auth.addr.%s.allow %s\n" % (subvolumes[0], self.auth_parameters))
mount_fd.write ("# By default insecure ports are not allowed.\n")
mount_fd.write ("# option rpc-auth.ports.insecure on\n")
mount_fd.write ("# option rpc-auth.ports.<volume>.insecure on\n")
mount_fd.write ("# By default all access is read-write.\n")
mount_fd.write ("# option nfs3.<volume>.volume-access read-only\n")
mount_fd.write ("# option nfs3.<volume>.volume-access read-only\n")
mount_fd.write ("# option nfs3.read-size 128Kb\n")
mount_fd.write ("# option nfs3.write-size 32Kb\n")
mount_fd.write ("# option nfs3.readdir-size 64Kb\n")
mount_fd.write ("# option nfs3.<volume>.read-size 64Kb\n")
mount_fd.write ("# option nfs3.<volume>.write-size 64Kb\n")
mount_fd.write ("# option nfs3.posix1.readdir-size 128Kb\n")
mount_fd.write ("end-volume\n\n")
return
if self.volume_size_client:
mount_fd.write ("volume quota\n")
mount_fd.write (" type features/quota\n")
mount_fd.write (" option disk-usage-limit %s\n" % self.volume_size_client)
mount_fd.write ("# option minimum-free-disk-limit 10GB\n"
"# minimum free disk value (default) 0\n")
mount_fd.write ("# option refresh-interval 10\n")
mount_fd.write (" subvolumes %s\n" % subvolumes[0])
mount_fd.write ("end-volume\n\n")
if self.enable_safe_mode:
return
self.performance_mode (subvolumes[0], mount_fd)
return
def performance_mode (self, cluster_subvolume, volfile_fd):
volfile_fd.write ("volume writebehind\n")
volfile_fd.write (" type performance/write-behind\n")
volfile_fd.write (" option cache-size 4MB\n")
volfile_fd.write ("# option enable-trickling-writes yes # Flush final write calls when network is free\n")
volfile_fd.write ("# option enable-O_SYNC yes # Enable O_SYNC for write-behind\n")
volfile_fd.write ("# option disable-for-first-nbytes 1 # Disable first nbytes with very small initial writes\n")
if self.volume_size_client:
volfile_fd.write (" subvolumes quota\n")
else:
volfile_fd.write (" subvolumes %s\n" % cluster_subvolume)
volfile_fd.write ("end-volume\n\n")
volfile_fd.write ("volume readahead\n")
volfile_fd.write (" type performance/read-ahead\n")
volfile_fd.write (" option page-count 4\n")
volfile_fd.write ("# option force-atime-update yes # force updating atimes, default off\n")
volfile_fd.write (" subvolumes writebehind\n")
volfile_fd.write ("end-volume\n\n")
volfile_fd.write ("volume iocache\n")
volfile_fd.write (" type performance/io-cache\n")
volfile_fd.write (" option cache-size %sMB\n" % cache_size)
volfile_fd.write (" option cache-timeout 1\n")
volfile_fd.write ("# option priority *.html:1,abc*:2 # Priority list for iocaching files\n")
volfile_fd.write (" subvolumes readahead\n")
volfile_fd.write ("end-volume\n\n")
volfile_fd.write ("volume quickread\n")
volfile_fd.write (" type performance/quick-read\n")
volfile_fd.write (" option cache-timeout 1\n")
volfile_fd.write (" option max-file-size 64kB\n")
volfile_fd.write (" subvolumes iocache\n")
volfile_fd.write ("end-volume\n\n")
volfile_fd.write ("volume statprefetch\n")
volfile_fd.write (" type performance/stat-prefetch\n")
volfile_fd.write (" subvolumes quickread\n")
volfile_fd.write ("end-volume\n\n")
return
def create_export_volfile (self):
cmdline = string.join (sys.argv, ' ')
i = 0
for export in self.host_dict[self.host]:
export_volfile = "%s/%s-%d.vol" % (self.conf_dir,
str(self.host + '-' + self.volume_name),
i)
i += 1
exp_fd = file ("%s" % (export_volfile),"w")
print "Generating server volfiles.. for server %s as '%s'" % (self.host,
export_volfile)
exp_fd.write ("## file auto generated by %s\n" % sys.argv[0])
exp_fd.write ("# Cmd line:\n")
exp_fd.write ("# $ %s\n\n" % cmdline)
exp_fd.write ("volume posix\n")
exp_fd.write (" type storage/posix\n")
exp_fd.write ("# option o-direct enable # (default: disable) boolean type only\n")
exp_fd.write ("# option export-statfs-size no # (default: yes) boolean type only\n")
exp_fd.write ("# option mandate-attribute off # (default: on) boolean type only\n")
exp_fd.write ("# option span-devices 8 # (default: 0) integer value\n")
exp_fd.write ("# option background-unlink yes # (default: no) boolean type\n")
exp_fd.write (" option directory %s\n" % export)
exp_fd.write ("end-volume\n\n")
if self.nfs:
exp_fd.write ("volume posix-ac\n")
exp_fd.write (" type features/access-control\n")
exp_fd.write (" subvolumes posix\n")
exp_fd.write ("end-volume\n\n")
if self.volume_size_server:
exp_fd.write ("volume quota\n")
exp_fd.write (" type features/quota\n")
exp_fd.write (" option disk-usage-limit %s\n" % self.volume_size_server)
exp_fd.write ("# option minimum-free-disk-limit 10GB"
" # minimum free disk value (default) 0\n")
exp_fd.write ("# option refresh-interval 10\n")
if self.nfs:
exp_fd.write (" subvolumes posix-ac\n")
else:
exp_fd.write (" subvolumes posix\n")
exp_fd.write ("end-volume\n\n")
exp_fd.write ("volume locks\n")
exp_fd.write (" type features/locks\n")
exp_fd.write ("# option mandatory on # Default off, used in specific applications\n")
if self.volume_size_server:
exp_fd.write (" subvolumes quota\n")
elif self.nfs:
exp_fd.write (" subvolumes posix-ac\n")
else:
exp_fd.write (" subvolumes posix\n")
exp_fd.write ("end-volume\n\n")
exp_fd.write ("volume replace-brick\n")
exp_fd.write (" type protocol/client\n")
if self.transport:
exp_fd.write (" option transport-type %s\n" % self.transport)
if self.gfs_port:
exp_fd.write (" option transport.remote-port %d\n" % self.gfs_port)
exp_fd.write (" option ping-timeout 42\n")
exp_fd.write ("end-volume\n\n")
exp_fd.write ("volume pump\n")
exp_fd.write (" type cluster/pump\n")
exp_fd.write (" subvolumes locks replace-brick\n")
exp_fd.write ("end-volume\n\n")
exp_fd.write ("volume %s\n" % export)
exp_fd.write (" type performance/io-threads\n")
exp_fd.write (" option thread-count 8\n")
exp_fd.write ("# option autoscaling yes # Heuristic for autoscaling threads on demand\n")
exp_fd.write ("# option min-threads 2 # min count for thread pool\n")
exp_fd.write ("# option max-threads 64 # max count for thread pool\n")
exp_fd.write (" subvolumes pump\n")
exp_fd.write ("end-volume\n\n")
for transport in self.transports:
exp_fd.write ("volume server-%s\n" % transport)
exp_fd.write (" type protocol/server\n")
exp_fd.write (" option transport-type %s\n" % transport)
exp_fd.write (" option auth.addr.%s.allow %s\n" %
(export, self.auth_parameters))
if transport == 'ib-verbs':
if self.gfs_ib_port:
exp_fd.write (" option listen-port %d\n" % self.gfs_ib_port)
exp_fd.write (" option transport.ib-verbs.port %d\n" %
self.ib_devport)
if transport == 'tcp':
if self.gfs_port:
exp_fd.write (" option listen-port %d\n" % self.gfs_port)
exp_fd.write (" option transport.socket.nodelay on\n")
exp_fd.write (" subvolumes %s\n" % export)
exp_fd.write ("end-volume\n\n")
return
|
