root / tools / burnin @ 88d31e5c
History | View | Annotate | Download (31.5 kB)
| 1 |
#!/usr/bin/python |
|---|---|
| 2 |
# |
| 3 |
|
| 4 |
# Copyright (C) 2006, 2007 Google Inc. |
| 5 |
# |
| 6 |
# This program is free software; you can redistribute it and/or modify |
| 7 |
# it under the terms of the GNU General Public License as published by |
| 8 |
# the Free Software Foundation; either version 2 of the License, or |
| 9 |
# (at your option) any later version. |
| 10 |
# |
| 11 |
# This program is distributed in the hope that it will be useful, but |
| 12 |
# WITHOUT ANY WARRANTY; without even the implied warranty of |
| 13 |
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 14 |
# General Public License for more details. |
| 15 |
# |
| 16 |
# You should have received a copy of the GNU General Public License |
| 17 |
# along with this program; if not, write to the Free Software |
| 18 |
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA |
| 19 |
# 02110-1301, USA. |
| 20 |
|
| 21 |
|
| 22 |
"""Burnin program |
| 23 |
|
| 24 |
""" |
| 25 |
|
| 26 |
import os |
| 27 |
import sys |
| 28 |
import optparse |
| 29 |
import time |
| 30 |
import socket |
| 31 |
import urllib |
| 32 |
from itertools import izip, islice, cycle |
| 33 |
from cStringIO import StringIO |
| 34 |
|
| 35 |
from ganeti import opcodes |
| 36 |
from ganeti import constants |
| 37 |
from ganeti import cli |
| 38 |
from ganeti import errors |
| 39 |
from ganeti import utils |
| 40 |
|
| 41 |
|
| 42 |
USAGE = ("\tburnin -o OS_NAME [options...] instance_name ...")
|
| 43 |
|
| 44 |
MAX_RETRIES = 3 |
| 45 |
|
| 46 |
class InstanceDown(Exception): |
| 47 |
"""The checked instance was not up""" |
| 48 |
|
| 49 |
|
| 50 |
class BurninFailure(Exception): |
| 51 |
"""Failure detected during burning""" |
| 52 |
|
| 53 |
|
| 54 |
def Usage(): |
| 55 |
"""Shows program usage information and exits the program.""" |
| 56 |
|
| 57 |
print >> sys.stderr, "Usage:" |
| 58 |
print >> sys.stderr, USAGE |
| 59 |
sys.exit(2) |
| 60 |
|
| 61 |
|
| 62 |
def Log(msg, indent=0): |
| 63 |
"""Simple function that prints out its argument. |
| 64 |
|
| 65 |
""" |
| 66 |
headers = {
|
| 67 |
0: "- ", |
| 68 |
1: "* ", |
| 69 |
2: "" |
| 70 |
} |
| 71 |
sys.stdout.write("%*s%s%s\n" % (2*indent, "",
|
| 72 |
headers.get(indent, " "), msg)) |
| 73 |
sys.stdout.flush() |
| 74 |
|
| 75 |
def Err(msg, exit_code=1): |
| 76 |
"""Simple error logging that prints to stderr. |
| 77 |
|
| 78 |
""" |
| 79 |
sys.stderr.write(msg + "\n") |
| 80 |
sys.stderr.flush() |
| 81 |
sys.exit(exit_code) |
| 82 |
|
| 83 |
|
| 84 |
class SimpleOpener(urllib.FancyURLopener): |
| 85 |
"""A simple url opener""" |
| 86 |
|
| 87 |
def prompt_user_passwd(self, host, realm, clear_cache = 0): |
| 88 |
"""No-interaction version of prompt_user_passwd.""" |
| 89 |
return None, None |
| 90 |
|
| 91 |
def http_error_default(self, url, fp, errcode, errmsg, headers): |
| 92 |
"""Custom error handling""" |
| 93 |
# make sure sockets are not left in CLOSE_WAIT, this is similar |
| 94 |
# but with a different exception to the BasicURLOpener class |
| 95 |
_ = fp.read() # throw away data |
| 96 |
fp.close() |
| 97 |
raise InstanceDown("HTTP error returned: code %s, msg %s" %
|
| 98 |
(errcode, errmsg)) |
| 99 |
|
| 100 |
|
| 101 |
class Burner(object): |
| 102 |
"""Burner class.""" |
| 103 |
|
| 104 |
def __init__(self): |
| 105 |
"""Constructor.""" |
| 106 |
utils.SetupLogging(constants.LOG_BURNIN, debug=False, stderr_logging=True) |
| 107 |
self.url_opener = SimpleOpener() |
| 108 |
self._feed_buf = StringIO() |
| 109 |
self.nodes = [] |
| 110 |
self.instances = [] |
| 111 |
self.to_rem = [] |
| 112 |
self.queued_ops = [] |
| 113 |
self.opts = None |
| 114 |
self.queue_retry = False |
| 115 |
self.disk_count = self.disk_growth = self.disk_size = None |
| 116 |
self.hvp = self.bep = None |
| 117 |
self.ParseOptions() |
| 118 |
self.cl = cli.GetClient() |
| 119 |
self.GetState() |
| 120 |
|
| 121 |
def ClearFeedbackBuf(self): |
| 122 |
"""Clear the feedback buffer.""" |
| 123 |
self._feed_buf.truncate(0) |
| 124 |
|
| 125 |
def GetFeedbackBuf(self): |
| 126 |
"""Return the contents of the buffer.""" |
| 127 |
return self._feed_buf.getvalue() |
| 128 |
|
| 129 |
def Feedback(self, msg): |
| 130 |
"""Acumulate feedback in our buffer.""" |
| 131 |
formatted_msg = "%s %s" % (time.ctime(utils.MergeTime(msg[0])), msg[2]) |
| 132 |
self._feed_buf.write(formatted_msg + "\n") |
| 133 |
if self.opts.verbose: |
| 134 |
Log(formatted_msg, indent=3) |
| 135 |
|
| 136 |
def MaybeRetry(self, retry_count, msg, fn, *args): |
| 137 |
"""Possibly retry a given function execution. |
| 138 |
|
| 139 |
@type retry_count: int |
| 140 |
@param retry_count: retry counter: |
| 141 |
- 0: non-retryable action |
| 142 |
- 1: last retry for a retryable action |
| 143 |
- MAX_RETRIES: original try for a retryable action |
| 144 |
@type msg: str |
| 145 |
@param msg: the kind of the operation |
| 146 |
@type fn: callable |
| 147 |
@param fn: the function to be called |
| 148 |
|
| 149 |
""" |
| 150 |
try: |
| 151 |
val = fn(*args) |
| 152 |
if retry_count > 0 and retry_count < MAX_RETRIES: |
| 153 |
Log("Idempotent %s succeeded after %d retries" %
|
| 154 |
(msg, MAX_RETRIES - retry_count)) |
| 155 |
return val |
| 156 |
except Exception, err: |
| 157 |
if retry_count == 0: |
| 158 |
Log("Non-idempotent %s failed, aborting" % (msg, ))
|
| 159 |
raise |
| 160 |
elif retry_count == 1: |
| 161 |
Log("Idempotent %s repeated failure, aborting" % (msg, ))
|
| 162 |
raise |
| 163 |
else: |
| 164 |
Log("Idempotent %s failed, retry #%d/%d: %s" %
|
| 165 |
(msg, MAX_RETRIES - retry_count + 1, MAX_RETRIES, err)) |
| 166 |
self.MaybeRetry(retry_count - 1, msg, fn, *args) |
| 167 |
|
| 168 |
def _ExecOp(self, *ops): |
| 169 |
"""Execute one or more opcodes and manage the exec buffer. |
| 170 |
|
| 171 |
@result: if only opcode has been passed, we return its result; |
| 172 |
otherwise we return the list of results |
| 173 |
|
| 174 |
""" |
| 175 |
job_id = cli.SendJob(ops, cl=self.cl) |
| 176 |
results = cli.PollJob(job_id, cl=self.cl, feedback_fn=self.Feedback) |
| 177 |
if len(ops) == 1: |
| 178 |
return results[0] |
| 179 |
else: |
| 180 |
return results |
| 181 |
|
| 182 |
def ExecOp(self, retry, *ops): |
| 183 |
"""Execute one or more opcodes and manage the exec buffer. |
| 184 |
|
| 185 |
@result: if only opcode has been passed, we return its result; |
| 186 |
otherwise we return the list of results |
| 187 |
|
| 188 |
""" |
| 189 |
if retry: |
| 190 |
rval = MAX_RETRIES |
| 191 |
else: |
| 192 |
rval = 0 |
| 193 |
return self.MaybeRetry(rval, "opcode", self._ExecOp, *ops) |
| 194 |
|
| 195 |
def ExecOrQueue(self, name, *ops): |
| 196 |
"""Execute an opcode and manage the exec buffer.""" |
| 197 |
if self.opts.parallel: |
| 198 |
self.queued_ops.append((ops, name)) |
| 199 |
else: |
| 200 |
return self.ExecOp(self.queue_retry, *ops) |
| 201 |
|
| 202 |
def StartBatch(self, retry): |
| 203 |
"""Start a new batch of jobs. |
| 204 |
|
| 205 |
@param retry: whether this is a retryable batch |
| 206 |
|
| 207 |
""" |
| 208 |
self.queued_ops = [] |
| 209 |
self.queue_retry = retry |
| 210 |
|
| 211 |
def CommitQueue(self): |
| 212 |
"""Execute all submitted opcodes in case of parallel burnin""" |
| 213 |
if not self.opts.parallel: |
| 214 |
return |
| 215 |
|
| 216 |
if self.queue_retry: |
| 217 |
rval = MAX_RETRIES |
| 218 |
else: |
| 219 |
rval = 0 |
| 220 |
|
| 221 |
try: |
| 222 |
results = self.MaybeRetry(rval, "jobset", self.ExecJobSet, |
| 223 |
self.queued_ops) |
| 224 |
finally: |
| 225 |
self.queued_ops = [] |
| 226 |
return results |
| 227 |
|
| 228 |
def ExecJobSet(self, jobs): |
| 229 |
"""Execute a set of jobs and return once all are done. |
| 230 |
|
| 231 |
The method will return the list of results, if all jobs are |
| 232 |
successful. Otherwise, OpExecError will be raised from within |
| 233 |
cli.py. |
| 234 |
|
| 235 |
""" |
| 236 |
self.ClearFeedbackBuf() |
| 237 |
job_ids = [cli.SendJob(row[0], cl=self.cl) for row in jobs] |
| 238 |
Log("Submitted job ID(s) %s" % ", ".join(job_ids), indent=1)
|
| 239 |
results = [] |
| 240 |
for jid, (_, iname) in zip(job_ids, jobs): |
| 241 |
Log("waiting for job %s for %s" % (jid, iname), indent=2)
|
| 242 |
try: |
| 243 |
results.append(cli.PollJob(jid, cl=self.cl, feedback_fn=self.Feedback)) |
| 244 |
except Exception, err: |
| 245 |
Log("Job for %s failed: %s" % (iname, err))
|
| 246 |
if len(results) != len(jobs): |
| 247 |
raise BurninFailure() |
| 248 |
return results |
| 249 |
|
| 250 |
def _DoCheckInstances(fn): |
| 251 |
"""Decorator for checking instances. |
| 252 |
|
| 253 |
""" |
| 254 |
def wrapper(self, *args, **kwargs): |
| 255 |
val = fn(self, *args, **kwargs) |
| 256 |
for instance in self.instances: |
| 257 |
self._CheckInstanceAlive(instance) |
| 258 |
return val |
| 259 |
|
| 260 |
return wrapper |
| 261 |
|
| 262 |
def _DoBatch(retry): |
| 263 |
"""Decorator for possible batch operations. |
| 264 |
|
| 265 |
Must come after the _DoCheckInstances decorator (if any). |
| 266 |
|
| 267 |
@param retry: whether this is a retryable batch, will be |
| 268 |
passed to StartBatch |
| 269 |
|
| 270 |
""" |
| 271 |
def wrap(fn): |
| 272 |
def batched(self, *args, **kwargs): |
| 273 |
self.StartBatch(retry) |
| 274 |
val = fn(self, *args, **kwargs) |
| 275 |
self.CommitQueue() |
| 276 |
return val |
| 277 |
return batched |
| 278 |
|
| 279 |
return wrap |
| 280 |
|
| 281 |
def ParseOptions(self): |
| 282 |
"""Parses the command line options. |
| 283 |
|
| 284 |
In case of command line errors, it will show the usage and exit the |
| 285 |
program. |
| 286 |
|
| 287 |
""" |
| 288 |
|
| 289 |
parser = optparse.OptionParser(usage="\n%s" % USAGE, |
| 290 |
version="%%prog (ganeti) %s" % |
| 291 |
constants.RELEASE_VERSION, |
| 292 |
option_class=cli.CliOption) |
| 293 |
|
| 294 |
parser.add_option("-o", "--os", dest="os", default=None,
|
| 295 |
help="OS to use during burnin", |
| 296 |
metavar="<OS>") |
| 297 |
parser.add_option("--disk-size", dest="disk_size",
|
| 298 |
help="Disk size (determines disk count)", |
| 299 |
default="128m", type="string", metavar="<size,size,...>") |
| 300 |
parser.add_option("--disk-growth", dest="disk_growth", help="Disk growth",
|
| 301 |
default="128m", type="string", metavar="<size,size,...>") |
| 302 |
parser.add_option("--mem-size", dest="mem_size", help="Memory size",
|
| 303 |
default=128, type="unit", metavar="<size>") |
| 304 |
parser.add_option("-v", "--verbose",
|
| 305 |
action="store_true", dest="verbose", default=False, |
| 306 |
help="print command execution messages to stdout") |
| 307 |
parser.add_option("--no-replace1", dest="do_replace1",
|
| 308 |
help="Skip disk replacement with the same secondary", |
| 309 |
action="store_false", default=True) |
| 310 |
parser.add_option("--no-replace2", dest="do_replace2",
|
| 311 |
help="Skip disk replacement with a different secondary", |
| 312 |
action="store_false", default=True) |
| 313 |
parser.add_option("--no-failover", dest="do_failover",
|
| 314 |
help="Skip instance failovers", action="store_false", |
| 315 |
default=True) |
| 316 |
parser.add_option("--no-migrate", dest="do_migrate",
|
| 317 |
help="Skip instance live migration", |
| 318 |
action="store_false", default=True) |
| 319 |
parser.add_option("--no-importexport", dest="do_importexport",
|
| 320 |
help="Skip instance export/import", action="store_false", |
| 321 |
default=True) |
| 322 |
parser.add_option("--no-startstop", dest="do_startstop",
|
| 323 |
help="Skip instance stop/start", action="store_false", |
| 324 |
default=True) |
| 325 |
parser.add_option("--no-reinstall", dest="do_reinstall",
|
| 326 |
help="Skip instance reinstall", action="store_false", |
| 327 |
default=True) |
| 328 |
parser.add_option("--no-reboot", dest="do_reboot",
|
| 329 |
help="Skip instance reboot", action="store_false", |
| 330 |
default=True) |
| 331 |
parser.add_option("--no-activate-disks", dest="do_activate_disks",
|
| 332 |
help="Skip disk activation/deactivation", |
| 333 |
action="store_false", default=True) |
| 334 |
parser.add_option("--no-add-disks", dest="do_addremove_disks",
|
| 335 |
help="Skip disk addition/removal", |
| 336 |
action="store_false", default=True) |
| 337 |
parser.add_option("--no-add-nics", dest="do_addremove_nics",
|
| 338 |
help="Skip NIC addition/removal", |
| 339 |
action="store_false", default=True) |
| 340 |
parser.add_option("--no-nics", dest="nics",
|
| 341 |
help="No network interfaces", action="store_const", |
| 342 |
const=[], default=[{}])
|
| 343 |
parser.add_option("--rename", dest="rename", default=None,
|
| 344 |
help="Give one unused instance name which is taken" |
| 345 |
" to start the renaming sequence", |
| 346 |
metavar="<instance_name>") |
| 347 |
parser.add_option("-t", "--disk-template", dest="disk_template",
|
| 348 |
choices=("diskless", "file", "plain", "drbd"),
|
| 349 |
default="drbd", |
| 350 |
help="Disk template (diskless, file, plain or drbd)" |
| 351 |
" [drbd]") |
| 352 |
parser.add_option("-n", "--nodes", dest="nodes", default="",
|
| 353 |
help="Comma separated list of nodes to perform" |
| 354 |
" the burnin on (defaults to all nodes)") |
| 355 |
parser.add_option("-I", "--iallocator", dest="iallocator",
|
| 356 |
default=None, type="string", |
| 357 |
help="Perform the allocation using an iallocator" |
| 358 |
" instead of fixed node spread (node restrictions no" |
| 359 |
" longer apply, therefore -n/--nodes must not be used") |
| 360 |
parser.add_option("-p", "--parallel", default=False, action="store_true",
|
| 361 |
dest="parallel", |
| 362 |
help="Enable parallelization of some operations in" |
| 363 |
" order to speed burnin or to test granular locking") |
| 364 |
parser.add_option("--net-timeout", default=15, type="int",
|
| 365 |
dest="net_timeout", |
| 366 |
help="The instance check network timeout in seconds" |
| 367 |
" (defaults to 15 seconds)") |
| 368 |
parser.add_option("-C", "--http-check", default=False, action="store_true",
|
| 369 |
dest="http_check", |
| 370 |
help="Enable checking of instance status via http," |
| 371 |
" looking for /hostname.txt that should contain the" |
| 372 |
" name of the instance") |
| 373 |
parser.add_option("-K", "--keep-instances", default=False,
|
| 374 |
action="store_true", |
| 375 |
dest="keep_instances", |
| 376 |
help="Leave instances on the cluster after burnin," |
| 377 |
" for investigation in case of errors or simply" |
| 378 |
" to use them") |
| 379 |
|
| 380 |
|
| 381 |
options, args = parser.parse_args() |
| 382 |
if len(args) < 1 or options.os is None: |
| 383 |
Usage() |
| 384 |
|
| 385 |
supported_disk_templates = (constants.DT_DISKLESS, |
| 386 |
constants.DT_FILE, |
| 387 |
constants.DT_PLAIN, |
| 388 |
constants.DT_DRBD8) |
| 389 |
if options.disk_template not in supported_disk_templates: |
| 390 |
Err("Unknown disk template '%s'" % options.disk_template)
|
| 391 |
|
| 392 |
if options.disk_template == constants.DT_DISKLESS: |
| 393 |
disk_size = disk_growth = [] |
| 394 |
options.do_addremove_disks = False |
| 395 |
else: |
| 396 |
disk_size = [utils.ParseUnit(v) for v in options.disk_size.split(",")]
|
| 397 |
disk_growth = [utils.ParseUnit(v) |
| 398 |
for v in options.disk_growth.split(",")]
|
| 399 |
if len(disk_growth) != len(disk_size): |
| 400 |
Err("Wrong disk sizes/growth combination")
|
| 401 |
if ((disk_size and options.disk_template == constants.DT_DISKLESS) or |
| 402 |
(not disk_size and options.disk_template != constants.DT_DISKLESS)): |
| 403 |
Err("Wrong disk count/disk template combination")
|
| 404 |
|
| 405 |
self.disk_size = disk_size |
| 406 |
self.disk_growth = disk_growth |
| 407 |
self.disk_count = len(disk_size) |
| 408 |
|
| 409 |
if options.nodes and options.iallocator: |
| 410 |
Err("Give either the nodes option or the iallocator option, not both")
|
| 411 |
|
| 412 |
self.opts = options |
| 413 |
self.instances = args |
| 414 |
self.bep = {
|
| 415 |
constants.BE_MEMORY: options.mem_size, |
| 416 |
constants.BE_VCPUS: 1, |
| 417 |
} |
| 418 |
self.hvp = {}
|
| 419 |
|
| 420 |
socket.setdefaulttimeout(options.net_timeout) |
| 421 |
|
| 422 |
def GetState(self): |
| 423 |
"""Read the cluster state from the config.""" |
| 424 |
if self.opts.nodes: |
| 425 |
names = self.opts.nodes.split(",")
|
| 426 |
else: |
| 427 |
names = [] |
| 428 |
try: |
| 429 |
op = opcodes.OpQueryNodes(output_fields=["name", "offline", "drained"], |
| 430 |
names=names, use_locking=True) |
| 431 |
result = self.ExecOp(True, op) |
| 432 |
except errors.GenericError, err: |
| 433 |
err_code, msg = cli.FormatError(err) |
| 434 |
Err(msg, exit_code=err_code) |
| 435 |
self.nodes = [data[0] for data in result if not (data[1] or data[2])] |
| 436 |
|
| 437 |
op_diagos = opcodes.OpDiagnoseOS(output_fields=["name", "valid"], names=[]) |
| 438 |
result = self.ExecOp(True, op_diagos) |
| 439 |
|
| 440 |
if not result: |
| 441 |
Err("Can't get the OS list")
|
| 442 |
|
| 443 |
# filter non-valid OS-es |
| 444 |
os_set = [val[0] for val in result if val[1]] |
| 445 |
|
| 446 |
if self.opts.os not in os_set: |
| 447 |
Err("OS '%s' not found" % self.opts.os)
|
| 448 |
|
| 449 |
@_DoCheckInstances |
| 450 |
@_DoBatch(False) |
| 451 |
def BurnCreateInstances(self): |
| 452 |
"""Create the given instances. |
| 453 |
|
| 454 |
""" |
| 455 |
self.to_rem = [] |
| 456 |
mytor = izip(cycle(self.nodes), |
| 457 |
islice(cycle(self.nodes), 1, None), |
| 458 |
self.instances) |
| 459 |
|
| 460 |
Log("Creating instances")
|
| 461 |
for pnode, snode, instance in mytor: |
| 462 |
Log("instance %s" % instance, indent=1)
|
| 463 |
if self.opts.iallocator: |
| 464 |
pnode = snode = None |
| 465 |
msg = "with iallocator %s" % self.opts.iallocator |
| 466 |
elif self.opts.disk_template not in constants.DTS_NET_MIRROR: |
| 467 |
snode = None |
| 468 |
msg = "on %s" % pnode |
| 469 |
else: |
| 470 |
msg = "on %s, %s" % (pnode, snode) |
| 471 |
|
| 472 |
Log(msg, indent=2) |
| 473 |
|
| 474 |
op = opcodes.OpCreateInstance(instance_name=instance, |
| 475 |
disks = [ {"size": size}
|
| 476 |
for size in self.disk_size], |
| 477 |
disk_template=self.opts.disk_template, |
| 478 |
nics=self.opts.nics, |
| 479 |
mode=constants.INSTANCE_CREATE, |
| 480 |
os_type=self.opts.os, |
| 481 |
pnode=pnode, |
| 482 |
snode=snode, |
| 483 |
start=True, |
| 484 |
ip_check=True, |
| 485 |
wait_for_sync=True, |
| 486 |
file_driver="loop", |
| 487 |
file_storage_dir=None, |
| 488 |
iallocator=self.opts.iallocator, |
| 489 |
beparams=self.bep, |
| 490 |
hvparams=self.hvp, |
| 491 |
) |
| 492 |
|
| 493 |
self.ExecOrQueue(instance, op) |
| 494 |
self.to_rem.append(instance) |
| 495 |
|
| 496 |
@_DoBatch(False) |
| 497 |
def BurnGrowDisks(self): |
| 498 |
"""Grow both the os and the swap disks by the requested amount, if any.""" |
| 499 |
Log("Growing disks")
|
| 500 |
for instance in self.instances: |
| 501 |
Log("instance %s" % instance, indent=1)
|
| 502 |
for idx, growth in enumerate(self.disk_growth): |
| 503 |
if growth > 0: |
| 504 |
op = opcodes.OpGrowDisk(instance_name=instance, disk=idx, |
| 505 |
amount=growth, wait_for_sync=True) |
| 506 |
Log("increase disk/%s by %s MB" % (idx, growth), indent=2)
|
| 507 |
self.ExecOrQueue(instance, op) |
| 508 |
|
| 509 |
@_DoBatch(True) |
| 510 |
def BurnReplaceDisks1D8(self): |
| 511 |
"""Replace disks on primary and secondary for drbd8.""" |
| 512 |
Log("Replacing disks on the same nodes")
|
| 513 |
for instance in self.instances: |
| 514 |
Log("instance %s" % instance, indent=1)
|
| 515 |
ops = [] |
| 516 |
for mode in constants.REPLACE_DISK_SEC, constants.REPLACE_DISK_PRI: |
| 517 |
op = opcodes.OpReplaceDisks(instance_name=instance, |
| 518 |
mode=mode, |
| 519 |
disks=[i for i in range(self.disk_count)]) |
| 520 |
Log("run %s" % mode, indent=2)
|
| 521 |
ops.append(op) |
| 522 |
self.ExecOrQueue(instance, *ops) |
| 523 |
|
| 524 |
@_DoBatch(True) |
| 525 |
def BurnReplaceDisks2(self): |
| 526 |
"""Replace secondary node.""" |
| 527 |
Log("Changing the secondary node")
|
| 528 |
mode = constants.REPLACE_DISK_CHG |
| 529 |
|
| 530 |
mytor = izip(islice(cycle(self.nodes), 2, None), |
| 531 |
self.instances) |
| 532 |
for tnode, instance in mytor: |
| 533 |
Log("instance %s" % instance, indent=1)
|
| 534 |
if self.opts.iallocator: |
| 535 |
tnode = None |
| 536 |
msg = "with iallocator %s" % self.opts.iallocator |
| 537 |
else: |
| 538 |
msg = tnode |
| 539 |
op = opcodes.OpReplaceDisks(instance_name=instance, |
| 540 |
mode=mode, |
| 541 |
remote_node=tnode, |
| 542 |
iallocator=self.opts.iallocator, |
| 543 |
disks=[i for i in range(self.disk_count)]) |
| 544 |
Log("run %s %s" % (mode, msg), indent=2)
|
| 545 |
self.ExecOrQueue(instance, op) |
| 546 |
|
| 547 |
@_DoCheckInstances |
| 548 |
@_DoBatch(False) |
| 549 |
def BurnFailover(self): |
| 550 |
"""Failover the instances.""" |
| 551 |
Log("Failing over instances")
|
| 552 |
for instance in self.instances: |
| 553 |
Log("instance %s" % instance, indent=1)
|
| 554 |
op = opcodes.OpFailoverInstance(instance_name=instance, |
| 555 |
ignore_consistency=False) |
| 556 |
|
| 557 |
self.ExecOrQueue(instance, op) |
| 558 |
|
| 559 |
@_DoBatch(False) |
| 560 |
def BurnMigrate(self): |
| 561 |
"""Migrate the instances.""" |
| 562 |
Log("Migrating instances")
|
| 563 |
for instance in self.instances: |
| 564 |
Log("instance %s" % instance, indent=1)
|
| 565 |
op1 = opcodes.OpMigrateInstance(instance_name=instance, live=True, |
| 566 |
cleanup=False) |
| 567 |
|
| 568 |
op2 = opcodes.OpMigrateInstance(instance_name=instance, live=True, |
| 569 |
cleanup=True) |
| 570 |
Log("migration and migration cleanup", indent=2)
|
| 571 |
self.ExecOrQueue(instance, op1, op2) |
| 572 |
|
| 573 |
@_DoCheckInstances |
| 574 |
@_DoBatch(False) |
| 575 |
def BurnImportExport(self): |
| 576 |
"""Export the instance, delete it, and import it back. |
| 577 |
|
| 578 |
""" |
| 579 |
Log("Exporting and re-importing instances")
|
| 580 |
mytor = izip(cycle(self.nodes), |
| 581 |
islice(cycle(self.nodes), 1, None), |
| 582 |
islice(cycle(self.nodes), 2, None), |
| 583 |
self.instances) |
| 584 |
|
| 585 |
for pnode, snode, enode, instance in mytor: |
| 586 |
Log("instance %s" % instance, indent=1)
|
| 587 |
# read the full name of the instance |
| 588 |
nam_op = opcodes.OpQueryInstances(output_fields=["name"], |
| 589 |
names=[instance], use_locking=True) |
| 590 |
full_name = self.ExecOp(False, nam_op)[0][0] |
| 591 |
|
| 592 |
if self.opts.iallocator: |
| 593 |
pnode = snode = None |
| 594 |
import_log_msg = ("import from %s"
|
| 595 |
" with iallocator %s" % |
| 596 |
(enode, self.opts.iallocator)) |
| 597 |
elif self.opts.disk_template not in constants.DTS_NET_MIRROR: |
| 598 |
snode = None |
| 599 |
import_log_msg = ("import from %s to %s" %
|
| 600 |
(enode, pnode)) |
| 601 |
else: |
| 602 |
import_log_msg = ("import from %s to %s, %s" %
|
| 603 |
(enode, pnode, snode)) |
| 604 |
|
| 605 |
exp_op = opcodes.OpExportInstance(instance_name=instance, |
| 606 |
target_node=enode, |
| 607 |
shutdown=True) |
| 608 |
rem_op = opcodes.OpRemoveInstance(instance_name=instance, |
| 609 |
ignore_failures=True) |
| 610 |
imp_dir = os.path.join(constants.EXPORT_DIR, full_name) |
| 611 |
imp_op = opcodes.OpCreateInstance(instance_name=instance, |
| 612 |
disks = [ {"size": size}
|
| 613 |
for size in self.disk_size], |
| 614 |
disk_template=self.opts.disk_template, |
| 615 |
nics=self.opts.nics, |
| 616 |
mode=constants.INSTANCE_IMPORT, |
| 617 |
src_node=enode, |
| 618 |
src_path=imp_dir, |
| 619 |
pnode=pnode, |
| 620 |
snode=snode, |
| 621 |
start=True, |
| 622 |
ip_check=True, |
| 623 |
wait_for_sync=True, |
| 624 |
file_storage_dir=None, |
| 625 |
file_driver="loop", |
| 626 |
iallocator=self.opts.iallocator, |
| 627 |
beparams=self.bep, |
| 628 |
hvparams=self.hvp, |
| 629 |
) |
| 630 |
|
| 631 |
erem_op = opcodes.OpRemoveExport(instance_name=instance) |
| 632 |
|
| 633 |
Log("export to node %s" % enode, indent=2)
|
| 634 |
Log("remove instance", indent=2)
|
| 635 |
Log(import_log_msg, indent=2) |
| 636 |
Log("remove export", indent=2)
|
| 637 |
self.ExecOrQueue(instance, exp_op, rem_op, imp_op, erem_op) |
| 638 |
|
| 639 |
def StopInstanceOp(self, instance): |
| 640 |
"""Stop given instance.""" |
| 641 |
return opcodes.OpShutdownInstance(instance_name=instance) |
| 642 |
|
| 643 |
def StartInstanceOp(self, instance): |
| 644 |
"""Start given instance.""" |
| 645 |
return opcodes.OpStartupInstance(instance_name=instance, force=False) |
| 646 |
|
| 647 |
def RenameInstanceOp(self, instance, instance_new): |
| 648 |
"""Rename instance.""" |
| 649 |
return opcodes.OpRenameInstance(instance_name=instance, |
| 650 |
new_name=instance_new) |
| 651 |
|
| 652 |
@_DoCheckInstances |
| 653 |
@_DoBatch(True) |
| 654 |
def BurnStopStart(self): |
| 655 |
"""Stop/start the instances.""" |
| 656 |
Log("Stopping and starting instances")
|
| 657 |
for instance in self.instances: |
| 658 |
Log("instance %s" % instance, indent=1)
|
| 659 |
op1 = self.StopInstanceOp(instance) |
| 660 |
op2 = self.StartInstanceOp(instance) |
| 661 |
self.ExecOrQueue(instance, op1, op2) |
| 662 |
|
| 663 |
@_DoBatch(False) |
| 664 |
def BurnRemove(self): |
| 665 |
"""Remove the instances.""" |
| 666 |
Log("Removing instances")
|
| 667 |
for instance in self.to_rem: |
| 668 |
Log("instance %s" % instance, indent=1)
|
| 669 |
op = opcodes.OpRemoveInstance(instance_name=instance, |
| 670 |
ignore_failures=True) |
| 671 |
self.ExecOrQueue(instance, op) |
| 672 |
|
| 673 |
def BurnRename(self): |
| 674 |
"""Rename the instances. |
| 675 |
|
| 676 |
Note that this function will not execute in parallel, since we |
| 677 |
only have one target for rename. |
| 678 |
|
| 679 |
""" |
| 680 |
Log("Renaming instances")
|
| 681 |
rename = self.opts.rename |
| 682 |
for instance in self.instances: |
| 683 |
Log("instance %s" % instance, indent=1)
|
| 684 |
op_stop1 = self.StopInstanceOp(instance) |
| 685 |
op_stop2 = self.StopInstanceOp(rename) |
| 686 |
op_rename1 = self.RenameInstanceOp(instance, rename) |
| 687 |
op_rename2 = self.RenameInstanceOp(rename, instance) |
| 688 |
op_start1 = self.StartInstanceOp(rename) |
| 689 |
op_start2 = self.StartInstanceOp(instance) |
| 690 |
self.ExecOp(False, op_stop1, op_rename1, op_start1) |
| 691 |
self._CheckInstanceAlive(rename) |
| 692 |
self.ExecOp(False, op_stop2, op_rename2, op_start2) |
| 693 |
self._CheckInstanceAlive(instance) |
| 694 |
|
| 695 |
@_DoCheckInstances |
| 696 |
@_DoBatch(True) |
| 697 |
def BurnReinstall(self): |
| 698 |
"""Reinstall the instances.""" |
| 699 |
Log("Reinstalling instances")
|
| 700 |
for instance in self.instances: |
| 701 |
Log("instance %s" % instance, indent=1)
|
| 702 |
op1 = self.StopInstanceOp(instance) |
| 703 |
op2 = opcodes.OpReinstallInstance(instance_name=instance) |
| 704 |
Log("reinstall without passing the OS", indent=2)
|
| 705 |
op3 = opcodes.OpReinstallInstance(instance_name=instance, |
| 706 |
os_type=self.opts.os) |
| 707 |
Log("reinstall specifying the OS", indent=2)
|
| 708 |
op4 = self.StartInstanceOp(instance) |
| 709 |
self.ExecOrQueue(instance, op1, op2, op3, op4) |
| 710 |
|
| 711 |
@_DoCheckInstances |
| 712 |
@_DoBatch(True) |
| 713 |
def BurnReboot(self): |
| 714 |
"""Reboot the instances.""" |
| 715 |
Log("Rebooting instances")
|
| 716 |
for instance in self.instances: |
| 717 |
Log("instance %s" % instance, indent=1)
|
| 718 |
ops = [] |
| 719 |
for reboot_type in constants.REBOOT_TYPES: |
| 720 |
op = opcodes.OpRebootInstance(instance_name=instance, |
| 721 |
reboot_type=reboot_type, |
| 722 |
ignore_secondaries=False) |
| 723 |
Log("reboot with type '%s'" % reboot_type, indent=2)
|
| 724 |
ops.append(op) |
| 725 |
self.ExecOrQueue(instance, *ops) |
| 726 |
|
| 727 |
@_DoCheckInstances |
| 728 |
@_DoBatch(True) |
| 729 |
def BurnActivateDisks(self): |
| 730 |
"""Activate and deactivate disks of the instances.""" |
| 731 |
Log("Activating/deactivating disks")
|
| 732 |
for instance in self.instances: |
| 733 |
Log("instance %s" % instance, indent=1)
|
| 734 |
op_start = self.StartInstanceOp(instance) |
| 735 |
op_act = opcodes.OpActivateInstanceDisks(instance_name=instance) |
| 736 |
op_deact = opcodes.OpDeactivateInstanceDisks(instance_name=instance) |
| 737 |
op_stop = self.StopInstanceOp(instance) |
| 738 |
Log("activate disks when online", indent=2)
|
| 739 |
Log("activate disks when offline", indent=2)
|
| 740 |
Log("deactivate disks (when offline)", indent=2)
|
| 741 |
self.ExecOrQueue(instance, op_act, op_stop, op_act, op_deact, op_start) |
| 742 |
|
| 743 |
@_DoCheckInstances |
| 744 |
@_DoBatch(False) |
| 745 |
def BurnAddRemoveDisks(self): |
| 746 |
"""Add and remove an extra disk for the instances.""" |
| 747 |
Log("Adding and removing disks")
|
| 748 |
for instance in self.instances: |
| 749 |
Log("instance %s" % instance, indent=1)
|
| 750 |
op_add = opcodes.OpSetInstanceParams(\ |
| 751 |
instance_name=instance, |
| 752 |
disks=[(constants.DDM_ADD, {"size": self.disk_size[0]})])
|
| 753 |
op_rem = opcodes.OpSetInstanceParams(\ |
| 754 |
instance_name=instance, disks=[(constants.DDM_REMOVE, {})])
|
| 755 |
op_stop = self.StopInstanceOp(instance) |
| 756 |
op_start = self.StartInstanceOp(instance) |
| 757 |
Log("adding a disk", indent=2)
|
| 758 |
Log("removing last disk", indent=2)
|
| 759 |
self.ExecOrQueue(instance, op_add, op_stop, op_rem, op_start) |
| 760 |
|
| 761 |
@_DoBatch(False) |
| 762 |
def BurnAddRemoveNICs(self): |
| 763 |
"""Add and remove an extra NIC for the instances.""" |
| 764 |
Log("Adding and removing NICs")
|
| 765 |
for instance in self.instances: |
| 766 |
Log("instance %s" % instance, indent=1)
|
| 767 |
op_add = opcodes.OpSetInstanceParams(\ |
| 768 |
instance_name=instance, nics=[(constants.DDM_ADD, {})])
|
| 769 |
op_rem = opcodes.OpSetInstanceParams(\ |
| 770 |
instance_name=instance, nics=[(constants.DDM_REMOVE, {})])
|
| 771 |
Log("adding a NIC", indent=2)
|
| 772 |
Log("removing last NIC", indent=2)
|
| 773 |
self.ExecOrQueue(instance, op_add, op_rem) |
| 774 |
|
| 775 |
def _CheckInstanceAlive(self, instance): |
| 776 |
"""Check if an instance is alive by doing http checks. |
| 777 |
|
| 778 |
This will try to retrieve the url on the instance /hostname.txt |
| 779 |
and check that it contains the hostname of the instance. In case |
| 780 |
we get ECONNREFUSED, we retry up to the net timeout seconds, for |
| 781 |
any other error we abort. |
| 782 |
|
| 783 |
""" |
| 784 |
if not self.opts.http_check: |
| 785 |
return |
| 786 |
end_time = time.time() + self.opts.net_timeout |
| 787 |
url = None |
| 788 |
while time.time() < end_time and url is None: |
| 789 |
try: |
| 790 |
url = self.url_opener.open("http://%s/hostname.txt" % instance)
|
| 791 |
except IOError: |
| 792 |
# here we can have connection refused, no route to host, etc. |
| 793 |
time.sleep(1) |
| 794 |
if url is None: |
| 795 |
raise InstanceDown(instance, "Cannot contact instance") |
| 796 |
hostname = url.read().strip() |
| 797 |
url.close() |
| 798 |
if hostname != instance: |
| 799 |
raise InstanceDown(instance, ("Hostname mismatch, expected %s, got %s" %
|
| 800 |
(instance, hostname))) |
| 801 |
|
| 802 |
def BurninCluster(self): |
| 803 |
"""Test a cluster intensively. |
| 804 |
|
| 805 |
This will create instances and then start/stop/failover them. |
| 806 |
It is safe for existing instances but could impact performance. |
| 807 |
|
| 808 |
""" |
| 809 |
|
| 810 |
opts = self.opts |
| 811 |
|
| 812 |
Log("Testing global parameters")
|
| 813 |
|
| 814 |
if (len(self.nodes) == 1 and |
| 815 |
opts.disk_template not in (constants.DT_DISKLESS, constants.DT_PLAIN, |
| 816 |
constants.DT_FILE)): |
| 817 |
Err("When one node is available/selected the disk template must"
|
| 818 |
" be 'diskless', 'file' or 'plain'") |
| 819 |
|
| 820 |
has_err = True |
| 821 |
try: |
| 822 |
self.BurnCreateInstances() |
| 823 |
if opts.do_replace1 and opts.disk_template in constants.DTS_NET_MIRROR: |
| 824 |
self.BurnReplaceDisks1D8() |
| 825 |
if (opts.do_replace2 and len(self.nodes) > 2 and |
| 826 |
opts.disk_template in constants.DTS_NET_MIRROR) : |
| 827 |
self.BurnReplaceDisks2() |
| 828 |
|
| 829 |
if (opts.disk_template != constants.DT_DISKLESS and |
| 830 |
utils.any(self.disk_growth, lambda n: n > 0)): |
| 831 |
self.BurnGrowDisks() |
| 832 |
|
| 833 |
if opts.do_failover and opts.disk_template in constants.DTS_NET_MIRROR: |
| 834 |
self.BurnFailover() |
| 835 |
|
| 836 |
if opts.do_migrate and opts.disk_template == constants.DT_DRBD8: |
| 837 |
self.BurnMigrate() |
| 838 |
|
| 839 |
if (opts.do_importexport and |
| 840 |
opts.disk_template not in (constants.DT_DISKLESS, |
| 841 |
constants.DT_FILE)): |
| 842 |
self.BurnImportExport() |
| 843 |
|
| 844 |
if opts.do_reinstall: |
| 845 |
self.BurnReinstall() |
| 846 |
|
| 847 |
if opts.do_reboot: |
| 848 |
self.BurnReboot() |
| 849 |
|
| 850 |
if opts.do_addremove_disks: |
| 851 |
self.BurnAddRemoveDisks() |
| 852 |
|
| 853 |
if opts.do_addremove_nics: |
| 854 |
self.BurnAddRemoveNICs() |
| 855 |
|
| 856 |
if opts.do_activate_disks: |
| 857 |
self.BurnActivateDisks() |
| 858 |
|
| 859 |
if opts.rename: |
| 860 |
self.BurnRename() |
| 861 |
|
| 862 |
if opts.do_startstop: |
| 863 |
self.BurnStopStart() |
| 864 |
|
| 865 |
has_err = False |
| 866 |
finally: |
| 867 |
if has_err: |
| 868 |
Log("Error detected: opcode buffer follows:\n\n")
|
| 869 |
Log(self.GetFeedbackBuf()) |
| 870 |
Log("\n\n")
|
| 871 |
if not self.opts.keep_instances: |
| 872 |
try: |
| 873 |
self.BurnRemove() |
| 874 |
except Exception, err: |
| 875 |
if has_err: # already detected errors, so errors in removal |
| 876 |
# are quite expected |
| 877 |
Log("Note: error detected during instance remove: %s" % str(err))
|
| 878 |
else: # non-expected error |
| 879 |
raise |
| 880 |
|
| 881 |
return 0 |
| 882 |
|
| 883 |
|
| 884 |
def main(): |
| 885 |
"""Main function""" |
| 886 |
|
| 887 |
burner = Burner() |
| 888 |
return burner.BurninCluster() |
| 889 |
|
| 890 |
|
| 891 |
if __name__ == "__main__": |
| 892 |
main() |