root / lib / cmdlib / instance_migration.py @ 87ed6b79
History | View | Annotate | Download (37.7 kB)
| 1 |
#
|
|---|---|
| 2 |
#
|
| 3 |
|
| 4 |
# Copyright (C) 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013 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 |
"""Logical units dealing with instance migration an failover."""
|
| 23 |
|
| 24 |
import logging |
| 25 |
import time |
| 26 |
|
| 27 |
from ganeti import constants |
| 28 |
from ganeti import errors |
| 29 |
from ganeti import locking |
| 30 |
from ganeti.masterd import iallocator |
| 31 |
from ganeti import utils |
| 32 |
from ganeti.cmdlib.base import LogicalUnit, Tasklet |
| 33 |
from ganeti.cmdlib.common import ExpandInstanceUuidAndName, \ |
| 34 |
CheckIAllocatorOrNode, ExpandNodeUuidAndName |
| 35 |
from ganeti.cmdlib.instance_storage import CheckDiskConsistency, \ |
| 36 |
ExpandCheckDisks, ShutdownInstanceDisks, AssembleInstanceDisks |
| 37 |
from ganeti.cmdlib.instance_utils import BuildInstanceHookEnvByObject, \ |
| 38 |
CheckTargetNodeIPolicy, ReleaseLocks, CheckNodeNotDrained, \ |
| 39 |
CopyLockList, CheckNodeFreeMemory, CheckInstanceBridgesExist |
| 40 |
|
| 41 |
import ganeti.masterd.instance |
| 42 |
|
| 43 |
|
| 44 |
def _ExpandNamesForMigration(lu): |
| 45 |
"""Expands names for use with L{TLMigrateInstance}.
|
| 46 |
|
| 47 |
@type lu: L{LogicalUnit}
|
| 48 |
|
| 49 |
"""
|
| 50 |
if lu.op.target_node is not None: |
| 51 |
(lu.op.target_node_uuid, lu.op.target_node) = \ |
| 52 |
ExpandNodeUuidAndName(lu.cfg, lu.op.target_node_uuid, lu.op.target_node) |
| 53 |
|
| 54 |
lu.needed_locks[locking.LEVEL_NODE] = [] |
| 55 |
lu.recalculate_locks[locking.LEVEL_NODE] = constants.LOCKS_REPLACE |
| 56 |
|
| 57 |
lu.needed_locks[locking.LEVEL_NODE_RES] = [] |
| 58 |
lu.recalculate_locks[locking.LEVEL_NODE_RES] = constants.LOCKS_REPLACE |
| 59 |
|
| 60 |
# The node allocation lock is actually only needed for externally replicated
|
| 61 |
# instances (e.g. sharedfile or RBD) and if an iallocator is used.
|
| 62 |
lu.needed_locks[locking.LEVEL_NODE_ALLOC] = [] |
| 63 |
|
| 64 |
|
| 65 |
def _DeclareLocksForMigration(lu, level): |
| 66 |
"""Declares locks for L{TLMigrateInstance}.
|
| 67 |
|
| 68 |
@type lu: L{LogicalUnit}
|
| 69 |
@param level: Lock level
|
| 70 |
|
| 71 |
"""
|
| 72 |
if level == locking.LEVEL_NODE_ALLOC:
|
| 73 |
assert lu.op.instance_name in lu.owned_locks(locking.LEVEL_INSTANCE) |
| 74 |
|
| 75 |
instance = lu.cfg.GetInstanceInfo(lu.op.instance_uuid) |
| 76 |
|
| 77 |
# Node locks are already declared here rather than at LEVEL_NODE as we need
|
| 78 |
# the instance object anyway to declare the node allocation lock.
|
| 79 |
if instance.disk_template in constants.DTS_EXT_MIRROR: |
| 80 |
if lu.op.target_node is None: |
| 81 |
lu.needed_locks[locking.LEVEL_NODE] = locking.ALL_SET |
| 82 |
lu.needed_locks[locking.LEVEL_NODE_ALLOC] = locking.ALL_SET |
| 83 |
else:
|
| 84 |
lu.needed_locks[locking.LEVEL_NODE] = [instance.primary_node, |
| 85 |
lu.op.target_node_uuid] |
| 86 |
del lu.recalculate_locks[locking.LEVEL_NODE]
|
| 87 |
else:
|
| 88 |
lu._LockInstancesNodes() # pylint: disable=W0212
|
| 89 |
|
| 90 |
elif level == locking.LEVEL_NODE:
|
| 91 |
# Node locks are declared together with the node allocation lock
|
| 92 |
assert (lu.needed_locks[locking.LEVEL_NODE] or |
| 93 |
lu.needed_locks[locking.LEVEL_NODE] is locking.ALL_SET)
|
| 94 |
|
| 95 |
elif level == locking.LEVEL_NODE_RES:
|
| 96 |
# Copy node locks
|
| 97 |
lu.needed_locks[locking.LEVEL_NODE_RES] = \ |
| 98 |
CopyLockList(lu.needed_locks[locking.LEVEL_NODE]) |
| 99 |
|
| 100 |
|
| 101 |
class LUInstanceFailover(LogicalUnit): |
| 102 |
"""Failover an instance.
|
| 103 |
|
| 104 |
"""
|
| 105 |
HPATH = "instance-failover"
|
| 106 |
HTYPE = constants.HTYPE_INSTANCE |
| 107 |
REQ_BGL = False
|
| 108 |
|
| 109 |
def CheckArguments(self): |
| 110 |
"""Check the arguments.
|
| 111 |
|
| 112 |
"""
|
| 113 |
self.iallocator = getattr(self.op, "iallocator", None) |
| 114 |
self.target_node = getattr(self.op, "target_node", None) |
| 115 |
|
| 116 |
def ExpandNames(self): |
| 117 |
self._ExpandAndLockInstance()
|
| 118 |
_ExpandNamesForMigration(self)
|
| 119 |
|
| 120 |
self._migrater = \
|
| 121 |
TLMigrateInstance(self, self.op.instance_uuid, self.op.instance_name, |
| 122 |
self.op.cleanup, True, False, |
| 123 |
self.op.ignore_consistency, True, |
| 124 |
self.op.shutdown_timeout, self.op.ignore_ipolicy) |
| 125 |
|
| 126 |
self.tasklets = [self._migrater] |
| 127 |
|
| 128 |
def DeclareLocks(self, level): |
| 129 |
_DeclareLocksForMigration(self, level)
|
| 130 |
|
| 131 |
def BuildHooksEnv(self): |
| 132 |
"""Build hooks env.
|
| 133 |
|
| 134 |
This runs on master, primary and secondary nodes of the instance.
|
| 135 |
|
| 136 |
"""
|
| 137 |
instance = self._migrater.instance
|
| 138 |
source_node_uuid = instance.primary_node |
| 139 |
target_node_uuid = self._migrater.target_node_uuid
|
| 140 |
env = {
|
| 141 |
"IGNORE_CONSISTENCY": self.op.ignore_consistency, |
| 142 |
"SHUTDOWN_TIMEOUT": self.op.shutdown_timeout, |
| 143 |
"OLD_PRIMARY": self.cfg.GetNodeName(source_node_uuid), |
| 144 |
"NEW_PRIMARY": self.cfg.GetNodeName(target_node_uuid), |
| 145 |
"FAILOVER_CLEANUP": self.op.cleanup, |
| 146 |
} |
| 147 |
|
| 148 |
if instance.disk_template in constants.DTS_INT_MIRROR: |
| 149 |
env["OLD_SECONDARY"] = self.cfg.GetNodeName(instance.secondary_nodes[0]) |
| 150 |
env["NEW_SECONDARY"] = self.cfg.GetNodeName(source_node_uuid) |
| 151 |
else:
|
| 152 |
env["OLD_SECONDARY"] = env["NEW_SECONDARY"] = "" |
| 153 |
|
| 154 |
env.update(BuildInstanceHookEnvByObject(self, instance))
|
| 155 |
|
| 156 |
return env
|
| 157 |
|
| 158 |
def BuildHooksNodes(self): |
| 159 |
"""Build hooks nodes.
|
| 160 |
|
| 161 |
"""
|
| 162 |
instance = self._migrater.instance
|
| 163 |
nl = [self.cfg.GetMasterNode()] + list(instance.secondary_nodes) |
| 164 |
nl.append(self._migrater.target_node_uuid)
|
| 165 |
return (nl, nl + [instance.primary_node])
|
| 166 |
|
| 167 |
|
| 168 |
class LUInstanceMigrate(LogicalUnit): |
| 169 |
"""Migrate an instance.
|
| 170 |
|
| 171 |
This is migration without shutting down, compared to the failover,
|
| 172 |
which is done with shutdown.
|
| 173 |
|
| 174 |
"""
|
| 175 |
HPATH = "instance-migrate"
|
| 176 |
HTYPE = constants.HTYPE_INSTANCE |
| 177 |
REQ_BGL = False
|
| 178 |
|
| 179 |
def ExpandNames(self): |
| 180 |
self._ExpandAndLockInstance()
|
| 181 |
_ExpandNamesForMigration(self)
|
| 182 |
|
| 183 |
self._migrater = \
|
| 184 |
TLMigrateInstance(self, self.op.instance_uuid, self.op.instance_name, |
| 185 |
self.op.cleanup, False, self.op.allow_failover, False, |
| 186 |
self.op.allow_runtime_changes,
|
| 187 |
constants.DEFAULT_SHUTDOWN_TIMEOUT, |
| 188 |
self.op.ignore_ipolicy)
|
| 189 |
|
| 190 |
self.tasklets = [self._migrater] |
| 191 |
|
| 192 |
def DeclareLocks(self, level): |
| 193 |
_DeclareLocksForMigration(self, level)
|
| 194 |
|
| 195 |
def BuildHooksEnv(self): |
| 196 |
"""Build hooks env.
|
| 197 |
|
| 198 |
This runs on master, primary and secondary nodes of the instance.
|
| 199 |
|
| 200 |
"""
|
| 201 |
instance = self._migrater.instance
|
| 202 |
source_node_uuid = instance.primary_node |
| 203 |
target_node_uuid = self._migrater.target_node_uuid
|
| 204 |
env = BuildInstanceHookEnvByObject(self, instance)
|
| 205 |
env.update({
|
| 206 |
"MIGRATE_LIVE": self._migrater.live, |
| 207 |
"MIGRATE_CLEANUP": self.op.cleanup, |
| 208 |
"OLD_PRIMARY": self.cfg.GetNodeName(source_node_uuid), |
| 209 |
"NEW_PRIMARY": self.cfg.GetNodeName(target_node_uuid), |
| 210 |
"ALLOW_RUNTIME_CHANGES": self.op.allow_runtime_changes, |
| 211 |
}) |
| 212 |
|
| 213 |
if instance.disk_template in constants.DTS_INT_MIRROR: |
| 214 |
env["OLD_SECONDARY"] = self.cfg.GetNodeName(instance.secondary_nodes[0]) |
| 215 |
env["NEW_SECONDARY"] = self.cfg.GetNodeName(source_node_uuid) |
| 216 |
else:
|
| 217 |
env["OLD_SECONDARY"] = env["NEW_SECONDARY"] = "" |
| 218 |
|
| 219 |
return env
|
| 220 |
|
| 221 |
def BuildHooksNodes(self): |
| 222 |
"""Build hooks nodes.
|
| 223 |
|
| 224 |
"""
|
| 225 |
instance = self._migrater.instance
|
| 226 |
snode_uuids = list(instance.secondary_nodes)
|
| 227 |
nl = [self.cfg.GetMasterNode(), instance.primary_node] + snode_uuids
|
| 228 |
nl.append(self._migrater.target_node_uuid)
|
| 229 |
return (nl, nl)
|
| 230 |
|
| 231 |
|
| 232 |
class TLMigrateInstance(Tasklet): |
| 233 |
"""Tasklet class for instance migration.
|
| 234 |
|
| 235 |
@type live: boolean
|
| 236 |
@ivar live: whether the migration will be done live or non-live;
|
| 237 |
this variable is initalized only after CheckPrereq has run
|
| 238 |
@type cleanup: boolean
|
| 239 |
@ivar cleanup: Wheater we cleanup from a failed migration
|
| 240 |
@type iallocator: string
|
| 241 |
@ivar iallocator: The iallocator used to determine target_node
|
| 242 |
@type target_node_uuid: string
|
| 243 |
@ivar target_node_uuid: If given, the target node UUID to reallocate the
|
| 244 |
instance to
|
| 245 |
@type failover: boolean
|
| 246 |
@ivar failover: Whether operation results in failover or migration
|
| 247 |
@type fallback: boolean
|
| 248 |
@ivar fallback: Whether fallback to failover is allowed if migration not
|
| 249 |
possible
|
| 250 |
@type ignore_consistency: boolean
|
| 251 |
@ivar ignore_consistency: Wheter we should ignore consistency between source
|
| 252 |
and target node
|
| 253 |
@type shutdown_timeout: int
|
| 254 |
@ivar shutdown_timeout: In case of failover timeout of the shutdown
|
| 255 |
@type ignore_ipolicy: bool
|
| 256 |
@ivar ignore_ipolicy: If true, we can ignore instance policy when migrating
|
| 257 |
|
| 258 |
"""
|
| 259 |
|
| 260 |
# Constants
|
| 261 |
_MIGRATION_POLL_INTERVAL = 1 # seconds |
| 262 |
_MIGRATION_FEEDBACK_INTERVAL = 10 # seconds |
| 263 |
|
| 264 |
def __init__(self, lu, instance_uuid, instance_name, cleanup, failover, |
| 265 |
fallback, ignore_consistency, allow_runtime_changes, |
| 266 |
shutdown_timeout, ignore_ipolicy): |
| 267 |
"""Initializes this class.
|
| 268 |
|
| 269 |
"""
|
| 270 |
Tasklet.__init__(self, lu)
|
| 271 |
|
| 272 |
# Parameters
|
| 273 |
self.instance_uuid = instance_uuid
|
| 274 |
self.instance_name = instance_name
|
| 275 |
self.cleanup = cleanup
|
| 276 |
self.live = False # will be overridden later |
| 277 |
self.failover = failover
|
| 278 |
self.fallback = fallback
|
| 279 |
self.ignore_consistency = ignore_consistency
|
| 280 |
self.shutdown_timeout = shutdown_timeout
|
| 281 |
self.ignore_ipolicy = ignore_ipolicy
|
| 282 |
self.allow_runtime_changes = allow_runtime_changes
|
| 283 |
|
| 284 |
def CheckPrereq(self): |
| 285 |
"""Check prerequisites.
|
| 286 |
|
| 287 |
This checks that the instance is in the cluster.
|
| 288 |
|
| 289 |
"""
|
| 290 |
(self.instance_uuid, self.instance_name) = \ |
| 291 |
ExpandInstanceUuidAndName(self.lu.cfg, self.instance_uuid, |
| 292 |
self.instance_name)
|
| 293 |
self.instance = self.cfg.GetInstanceInfo(self.instance_uuid) |
| 294 |
assert self.instance is not None |
| 295 |
cluster = self.cfg.GetClusterInfo()
|
| 296 |
|
| 297 |
if (not self.cleanup and |
| 298 |
not self.instance.admin_state == constants.ADMINST_UP and |
| 299 |
not self.failover and self.fallback): |
| 300 |
self.lu.LogInfo("Instance is marked down or offline, fallback allowed," |
| 301 |
" switching to failover")
|
| 302 |
self.failover = True |
| 303 |
|
| 304 |
if self.instance.disk_template not in constants.DTS_MIRRORED: |
| 305 |
if self.failover: |
| 306 |
text = "failovers"
|
| 307 |
else:
|
| 308 |
text = "migrations"
|
| 309 |
raise errors.OpPrereqError("Instance's disk layout '%s' does not allow" |
| 310 |
" %s" % (self.instance.disk_template, text), |
| 311 |
errors.ECODE_STATE) |
| 312 |
|
| 313 |
if self.instance.disk_template in constants.DTS_EXT_MIRROR: |
| 314 |
CheckIAllocatorOrNode(self.lu, "iallocator", "target_node") |
| 315 |
|
| 316 |
if self.lu.op.iallocator: |
| 317 |
assert locking.NAL in self.lu.owned_locks(locking.LEVEL_NODE_ALLOC) |
| 318 |
self._RunAllocator()
|
| 319 |
else:
|
| 320 |
# We set set self.target_node_uuid as it is required by
|
| 321 |
# BuildHooksEnv
|
| 322 |
self.target_node_uuid = self.lu.op.target_node_uuid |
| 323 |
|
| 324 |
# Check that the target node is correct in terms of instance policy
|
| 325 |
nodeinfo = self.cfg.GetNodeInfo(self.target_node_uuid) |
| 326 |
group_info = self.cfg.GetNodeGroup(nodeinfo.group)
|
| 327 |
ipolicy = ganeti.masterd.instance.CalculateGroupIPolicy(cluster, |
| 328 |
group_info) |
| 329 |
CheckTargetNodeIPolicy(self.lu, ipolicy, self.instance, nodeinfo, |
| 330 |
self.cfg, ignore=self.ignore_ipolicy) |
| 331 |
|
| 332 |
# self.target_node is already populated, either directly or by the
|
| 333 |
# iallocator run
|
| 334 |
target_node_uuid = self.target_node_uuid
|
| 335 |
if self.target_node_uuid == self.instance.primary_node: |
| 336 |
raise errors.OpPrereqError(
|
| 337 |
"Cannot migrate instance %s to its primary (%s)" %
|
| 338 |
(self.instance.name,
|
| 339 |
self.cfg.GetNodeName(self.instance.primary_node)), |
| 340 |
errors.ECODE_STATE) |
| 341 |
|
| 342 |
if len(self.lu.tasklets) == 1: |
| 343 |
# It is safe to release locks only when we're the only tasklet
|
| 344 |
# in the LU
|
| 345 |
ReleaseLocks(self.lu, locking.LEVEL_NODE,
|
| 346 |
keep=[self.instance.primary_node, self.target_node_uuid]) |
| 347 |
ReleaseLocks(self.lu, locking.LEVEL_NODE_ALLOC)
|
| 348 |
|
| 349 |
else:
|
| 350 |
secondary_node_uuids = self.instance.secondary_nodes
|
| 351 |
if not secondary_node_uuids: |
| 352 |
raise errors.ConfigurationError("No secondary node but using" |
| 353 |
" %s disk template" %
|
| 354 |
self.instance.disk_template)
|
| 355 |
self.target_node_uuid = target_node_uuid = secondary_node_uuids[0] |
| 356 |
if self.lu.op.iallocator or \ |
| 357 |
(self.lu.op.target_node_uuid and |
| 358 |
self.lu.op.target_node_uuid != target_node_uuid):
|
| 359 |
if self.failover: |
| 360 |
text = "failed over"
|
| 361 |
else:
|
| 362 |
text = "migrated"
|
| 363 |
raise errors.OpPrereqError("Instances with disk template %s cannot" |
| 364 |
" be %s to arbitrary nodes"
|
| 365 |
" (neither an iallocator nor a target"
|
| 366 |
" node can be passed)" %
|
| 367 |
(self.instance.disk_template, text),
|
| 368 |
errors.ECODE_INVAL) |
| 369 |
nodeinfo = self.cfg.GetNodeInfo(target_node_uuid)
|
| 370 |
group_info = self.cfg.GetNodeGroup(nodeinfo.group)
|
| 371 |
ipolicy = ganeti.masterd.instance.CalculateGroupIPolicy(cluster, |
| 372 |
group_info) |
| 373 |
CheckTargetNodeIPolicy(self.lu, ipolicy, self.instance, nodeinfo, |
| 374 |
self.cfg, ignore=self.ignore_ipolicy) |
| 375 |
|
| 376 |
i_be = cluster.FillBE(self.instance)
|
| 377 |
|
| 378 |
# check memory requirements on the secondary node
|
| 379 |
if (not self.cleanup and |
| 380 |
(not self.failover or |
| 381 |
self.instance.admin_state == constants.ADMINST_UP)):
|
| 382 |
self.tgt_free_mem = CheckNodeFreeMemory(
|
| 383 |
self.lu, target_node_uuid,
|
| 384 |
"migrating instance %s" % self.instance.name, |
| 385 |
i_be[constants.BE_MINMEM], self.instance.hypervisor,
|
| 386 |
self.cfg.GetClusterInfo().hvparams[self.instance.hypervisor]) |
| 387 |
else:
|
| 388 |
self.lu.LogInfo("Not checking memory on the secondary node as" |
| 389 |
" instance will not be started")
|
| 390 |
|
| 391 |
# check if failover must be forced instead of migration
|
| 392 |
if (not self.cleanup and not self.failover and |
| 393 |
i_be[constants.BE_ALWAYS_FAILOVER]): |
| 394 |
self.lu.LogInfo("Instance configured to always failover; fallback" |
| 395 |
" to failover")
|
| 396 |
self.failover = True |
| 397 |
|
| 398 |
# check bridge existance
|
| 399 |
CheckInstanceBridgesExist(self.lu, self.instance, |
| 400 |
node_uuid=target_node_uuid) |
| 401 |
|
| 402 |
if not self.cleanup: |
| 403 |
CheckNodeNotDrained(self.lu, target_node_uuid)
|
| 404 |
if not self.failover: |
| 405 |
result = self.rpc.call_instance_migratable(self.instance.primary_node, |
| 406 |
self.instance)
|
| 407 |
if result.fail_msg and self.fallback: |
| 408 |
self.lu.LogInfo("Can't migrate, instance offline, fallback to" |
| 409 |
" failover")
|
| 410 |
self.failover = True |
| 411 |
else:
|
| 412 |
result.Raise("Can't migrate, please use failover",
|
| 413 |
prereq=True, ecode=errors.ECODE_STATE)
|
| 414 |
|
| 415 |
assert not (self.failover and self.cleanup) |
| 416 |
|
| 417 |
if not self.failover: |
| 418 |
if self.lu.op.live is not None and self.lu.op.mode is not None: |
| 419 |
raise errors.OpPrereqError("Only one of the 'live' and 'mode'" |
| 420 |
" parameters are accepted",
|
| 421 |
errors.ECODE_INVAL) |
| 422 |
if self.lu.op.live is not None: |
| 423 |
if self.lu.op.live: |
| 424 |
self.lu.op.mode = constants.HT_MIGRATION_LIVE
|
| 425 |
else:
|
| 426 |
self.lu.op.mode = constants.HT_MIGRATION_NONLIVE
|
| 427 |
# reset the 'live' parameter to None so that repeated
|
| 428 |
# invocations of CheckPrereq do not raise an exception
|
| 429 |
self.lu.op.live = None |
| 430 |
elif self.lu.op.mode is None: |
| 431 |
# read the default value from the hypervisor
|
| 432 |
i_hv = cluster.FillHV(self.instance, skip_globals=False) |
| 433 |
self.lu.op.mode = i_hv[constants.HV_MIGRATION_MODE]
|
| 434 |
|
| 435 |
self.live = self.lu.op.mode == constants.HT_MIGRATION_LIVE |
| 436 |
else:
|
| 437 |
# Failover is never live
|
| 438 |
self.live = False |
| 439 |
|
| 440 |
if not (self.failover or self.cleanup): |
| 441 |
remote_info = self.rpc.call_instance_info(
|
| 442 |
self.instance.primary_node, self.instance.name, |
| 443 |
self.instance.hypervisor, cluster.hvparams[self.instance.hypervisor]) |
| 444 |
remote_info.Raise("Error checking instance on node %s" %
|
| 445 |
self.cfg.GetNodeName(self.instance.primary_node)) |
| 446 |
instance_running = bool(remote_info.payload)
|
| 447 |
if instance_running:
|
| 448 |
self.current_mem = int(remote_info.payload["memory"]) |
| 449 |
|
| 450 |
def _RunAllocator(self): |
| 451 |
"""Run the allocator based on input opcode.
|
| 452 |
|
| 453 |
"""
|
| 454 |
assert locking.NAL in self.lu.owned_locks(locking.LEVEL_NODE_ALLOC) |
| 455 |
|
| 456 |
# FIXME: add a self.ignore_ipolicy option
|
| 457 |
req = iallocator.IAReqRelocate( |
| 458 |
inst_uuid=self.instance_uuid,
|
| 459 |
relocate_from_node_uuids=[self.instance.primary_node])
|
| 460 |
ial = iallocator.IAllocator(self.cfg, self.rpc, req) |
| 461 |
|
| 462 |
ial.Run(self.lu.op.iallocator)
|
| 463 |
|
| 464 |
if not ial.success: |
| 465 |
raise errors.OpPrereqError("Can't compute nodes using" |
| 466 |
" iallocator '%s': %s" %
|
| 467 |
(self.lu.op.iallocator, ial.info),
|
| 468 |
errors.ECODE_NORES) |
| 469 |
self.target_node_uuid = self.cfg.GetNodeInfoByName(ial.result[0]).uuid |
| 470 |
self.lu.LogInfo("Selected nodes for instance %s via iallocator %s: %s", |
| 471 |
self.instance_name, self.lu.op.iallocator, |
| 472 |
utils.CommaJoin(ial.result)) |
| 473 |
|
| 474 |
def _WaitUntilSync(self): |
| 475 |
"""Poll with custom rpc for disk sync.
|
| 476 |
|
| 477 |
This uses our own step-based rpc call.
|
| 478 |
|
| 479 |
"""
|
| 480 |
self.feedback_fn("* wait until resync is done") |
| 481 |
all_done = False
|
| 482 |
while not all_done: |
| 483 |
all_done = True
|
| 484 |
result = self.rpc.call_drbd_wait_sync(self.all_node_uuids, |
| 485 |
(self.instance.disks,
|
| 486 |
self.instance))
|
| 487 |
min_percent = 100
|
| 488 |
for node_uuid, nres in result.items(): |
| 489 |
nres.Raise("Cannot resync disks on node %s" %
|
| 490 |
self.cfg.GetNodeName(node_uuid))
|
| 491 |
node_done, node_percent = nres.payload |
| 492 |
all_done = all_done and node_done
|
| 493 |
if node_percent is not None: |
| 494 |
min_percent = min(min_percent, node_percent)
|
| 495 |
if not all_done: |
| 496 |
if min_percent < 100: |
| 497 |
self.feedback_fn(" - progress: %.1f%%" % min_percent) |
| 498 |
time.sleep(2)
|
| 499 |
|
| 500 |
def _EnsureSecondary(self, node_uuid): |
| 501 |
"""Demote a node to secondary.
|
| 502 |
|
| 503 |
"""
|
| 504 |
self.feedback_fn("* switching node %s to secondary mode" % |
| 505 |
self.cfg.GetNodeName(node_uuid))
|
| 506 |
|
| 507 |
result = self.rpc.call_blockdev_close(node_uuid, self.instance.name, |
| 508 |
(self.instance.disks, self.instance)) |
| 509 |
result.Raise("Cannot change disk to secondary on node %s" %
|
| 510 |
self.cfg.GetNodeName(node_uuid))
|
| 511 |
|
| 512 |
def _GoStandalone(self): |
| 513 |
"""Disconnect from the network.
|
| 514 |
|
| 515 |
"""
|
| 516 |
self.feedback_fn("* changing into standalone mode") |
| 517 |
result = self.rpc.call_drbd_disconnect_net(
|
| 518 |
self.all_node_uuids, (self.instance.disks, self.instance)) |
| 519 |
for node_uuid, nres in result.items(): |
| 520 |
nres.Raise("Cannot disconnect disks node %s" %
|
| 521 |
self.cfg.GetNodeName(node_uuid))
|
| 522 |
|
| 523 |
def _GoReconnect(self, multimaster): |
| 524 |
"""Reconnect to the network.
|
| 525 |
|
| 526 |
"""
|
| 527 |
if multimaster:
|
| 528 |
msg = "dual-master"
|
| 529 |
else:
|
| 530 |
msg = "single-master"
|
| 531 |
self.feedback_fn("* changing disks into %s mode" % msg) |
| 532 |
result = self.rpc.call_drbd_attach_net(self.all_node_uuids, |
| 533 |
(self.instance.disks, self.instance), |
| 534 |
self.instance.name, multimaster)
|
| 535 |
for node_uuid, nres in result.items(): |
| 536 |
nres.Raise("Cannot change disks config on node %s" %
|
| 537 |
self.cfg.GetNodeName(node_uuid))
|
| 538 |
|
| 539 |
def _ExecCleanup(self): |
| 540 |
"""Try to cleanup after a failed migration.
|
| 541 |
|
| 542 |
The cleanup is done by:
|
| 543 |
- check that the instance is running only on one node
|
| 544 |
(and update the config if needed)
|
| 545 |
- change disks on its secondary node to secondary
|
| 546 |
- wait until disks are fully synchronized
|
| 547 |
- disconnect from the network
|
| 548 |
- change disks into single-master mode
|
| 549 |
- wait again until disks are fully synchronized
|
| 550 |
|
| 551 |
"""
|
| 552 |
# check running on only one node
|
| 553 |
self.feedback_fn("* checking where the instance actually runs" |
| 554 |
" (if this hangs, the hypervisor might be in"
|
| 555 |
" a bad state)")
|
| 556 |
cluster_hvparams = self.cfg.GetClusterInfo().hvparams
|
| 557 |
ins_l = self.rpc.call_instance_list(self.all_node_uuids, |
| 558 |
[self.instance.hypervisor],
|
| 559 |
cluster_hvparams) |
| 560 |
for node_uuid, result in ins_l.items(): |
| 561 |
result.Raise("Can't contact node %s" % node_uuid)
|
| 562 |
|
| 563 |
runningon_source = self.instance.name in \ |
| 564 |
ins_l[self.source_node_uuid].payload
|
| 565 |
runningon_target = self.instance.name in \ |
| 566 |
ins_l[self.target_node_uuid].payload
|
| 567 |
|
| 568 |
if runningon_source and runningon_target: |
| 569 |
raise errors.OpExecError("Instance seems to be running on two nodes," |
| 570 |
" or the hypervisor is confused; you will have"
|
| 571 |
" to ensure manually that it runs only on one"
|
| 572 |
" and restart this operation")
|
| 573 |
|
| 574 |
if not (runningon_source or runningon_target): |
| 575 |
raise errors.OpExecError("Instance does not seem to be running at all;" |
| 576 |
" in this case it's safer to repair by"
|
| 577 |
" running 'gnt-instance stop' to ensure disk"
|
| 578 |
" shutdown, and then restarting it")
|
| 579 |
|
| 580 |
if runningon_target:
|
| 581 |
# the migration has actually succeeded, we need to update the config
|
| 582 |
self.feedback_fn("* instance running on secondary node (%s)," |
| 583 |
" updating config" %
|
| 584 |
self.cfg.GetNodeName(self.target_node_uuid)) |
| 585 |
self.instance.primary_node = self.target_node_uuid |
| 586 |
self.cfg.Update(self.instance, self.feedback_fn) |
| 587 |
demoted_node_uuid = self.source_node_uuid
|
| 588 |
else:
|
| 589 |
self.feedback_fn("* instance confirmed to be running on its" |
| 590 |
" primary node (%s)" %
|
| 591 |
self.cfg.GetNodeName(self.source_node_uuid)) |
| 592 |
demoted_node_uuid = self.target_node_uuid
|
| 593 |
|
| 594 |
if self.instance.disk_template in constants.DTS_INT_MIRROR: |
| 595 |
self._EnsureSecondary(demoted_node_uuid)
|
| 596 |
try:
|
| 597 |
self._WaitUntilSync()
|
| 598 |
except errors.OpExecError:
|
| 599 |
# we ignore here errors, since if the device is standalone, it
|
| 600 |
# won't be able to sync
|
| 601 |
pass
|
| 602 |
self._GoStandalone()
|
| 603 |
self._GoReconnect(False) |
| 604 |
self._WaitUntilSync()
|
| 605 |
|
| 606 |
self.feedback_fn("* done") |
| 607 |
|
| 608 |
def _RevertDiskStatus(self): |
| 609 |
"""Try to revert the disk status after a failed migration.
|
| 610 |
|
| 611 |
"""
|
| 612 |
if self.instance.disk_template in constants.DTS_EXT_MIRROR: |
| 613 |
return
|
| 614 |
|
| 615 |
try:
|
| 616 |
self._EnsureSecondary(self.target_node_uuid) |
| 617 |
self._GoStandalone()
|
| 618 |
self._GoReconnect(False) |
| 619 |
self._WaitUntilSync()
|
| 620 |
except errors.OpExecError, err:
|
| 621 |
self.lu.LogWarning("Migration failed and I can't reconnect the drives," |
| 622 |
" please try to recover the instance manually;"
|
| 623 |
" error '%s'" % str(err)) |
| 624 |
|
| 625 |
def _AbortMigration(self): |
| 626 |
"""Call the hypervisor code to abort a started migration.
|
| 627 |
|
| 628 |
"""
|
| 629 |
abort_result = self.rpc.call_instance_finalize_migration_dst(
|
| 630 |
self.target_node_uuid, self.instance, self.migration_info, |
| 631 |
False)
|
| 632 |
abort_msg = abort_result.fail_msg |
| 633 |
if abort_msg:
|
| 634 |
logging.error("Aborting migration failed on target node %s: %s",
|
| 635 |
self.cfg.GetNodeName(self.target_node_uuid), abort_msg) |
| 636 |
# Don't raise an exception here, as we stil have to try to revert the
|
| 637 |
# disk status, even if this step failed.
|
| 638 |
|
| 639 |
abort_result = self.rpc.call_instance_finalize_migration_src(
|
| 640 |
self.source_node_uuid, self.instance, False, self.live) |
| 641 |
abort_msg = abort_result.fail_msg |
| 642 |
if abort_msg:
|
| 643 |
logging.error("Aborting migration failed on source node %s: %s",
|
| 644 |
self.cfg.GetNodeName(self.source_node_uuid), abort_msg) |
| 645 |
|
| 646 |
def _ExecMigration(self): |
| 647 |
"""Migrate an instance.
|
| 648 |
|
| 649 |
The migrate is done by:
|
| 650 |
- change the disks into dual-master mode
|
| 651 |
- wait until disks are fully synchronized again
|
| 652 |
- migrate the instance
|
| 653 |
- change disks on the new secondary node (the old primary) to secondary
|
| 654 |
- wait until disks are fully synchronized
|
| 655 |
- change disks into single-master mode
|
| 656 |
|
| 657 |
"""
|
| 658 |
# Check for hypervisor version mismatch and warn the user.
|
| 659 |
hvspecs = [(self.instance.hypervisor,
|
| 660 |
self.cfg.GetClusterInfo().hvparams[self.instance.hypervisor])] |
| 661 |
nodeinfo = self.rpc.call_node_info(
|
| 662 |
[self.source_node_uuid, self.target_node_uuid], None, hvspecs) |
| 663 |
for ninfo in nodeinfo.values(): |
| 664 |
ninfo.Raise("Unable to retrieve node information from node '%s'" %
|
| 665 |
ninfo.node) |
| 666 |
(_, _, (src_info, )) = nodeinfo[self.source_node_uuid].payload
|
| 667 |
(_, _, (dst_info, )) = nodeinfo[self.target_node_uuid].payload
|
| 668 |
|
| 669 |
if ((constants.HV_NODEINFO_KEY_VERSION in src_info) and |
| 670 |
(constants.HV_NODEINFO_KEY_VERSION in dst_info)):
|
| 671 |
src_version = src_info[constants.HV_NODEINFO_KEY_VERSION] |
| 672 |
dst_version = dst_info[constants.HV_NODEINFO_KEY_VERSION] |
| 673 |
if src_version != dst_version:
|
| 674 |
self.feedback_fn("* warning: hypervisor version mismatch between" |
| 675 |
" source (%s) and target (%s) node" %
|
| 676 |
(src_version, dst_version)) |
| 677 |
|
| 678 |
self.feedback_fn("* checking disk consistency between source and target") |
| 679 |
for (idx, dev) in enumerate(self.instance.disks): |
| 680 |
if not CheckDiskConsistency(self.lu, self.instance, dev, |
| 681 |
self.target_node_uuid,
|
| 682 |
False):
|
| 683 |
raise errors.OpExecError("Disk %s is degraded or not fully" |
| 684 |
" synchronized on target node,"
|
| 685 |
" aborting migration" % idx)
|
| 686 |
|
| 687 |
if self.current_mem > self.tgt_free_mem: |
| 688 |
if not self.allow_runtime_changes: |
| 689 |
raise errors.OpExecError("Memory ballooning not allowed and not enough" |
| 690 |
" free memory to fit instance %s on target"
|
| 691 |
" node %s (have %dMB, need %dMB)" %
|
| 692 |
(self.instance.name,
|
| 693 |
self.cfg.GetNodeName(self.target_node_uuid), |
| 694 |
self.tgt_free_mem, self.current_mem)) |
| 695 |
self.feedback_fn("* setting instance memory to %s" % self.tgt_free_mem) |
| 696 |
rpcres = self.rpc.call_instance_balloon_memory(self.instance.primary_node, |
| 697 |
self.instance,
|
| 698 |
self.tgt_free_mem)
|
| 699 |
rpcres.Raise("Cannot modify instance runtime memory")
|
| 700 |
|
| 701 |
# First get the migration information from the remote node
|
| 702 |
result = self.rpc.call_migration_info(self.source_node_uuid, self.instance) |
| 703 |
msg = result.fail_msg |
| 704 |
if msg:
|
| 705 |
log_err = ("Failed fetching source migration information from %s: %s" %
|
| 706 |
(self.cfg.GetNodeName(self.source_node_uuid), msg)) |
| 707 |
logging.error(log_err) |
| 708 |
raise errors.OpExecError(log_err)
|
| 709 |
|
| 710 |
self.migration_info = migration_info = result.payload
|
| 711 |
|
| 712 |
if self.instance.disk_template not in constants.DTS_EXT_MIRROR: |
| 713 |
# Then switch the disks to master/master mode
|
| 714 |
self._EnsureSecondary(self.target_node_uuid) |
| 715 |
self._GoStandalone()
|
| 716 |
self._GoReconnect(True) |
| 717 |
self._WaitUntilSync()
|
| 718 |
|
| 719 |
self.feedback_fn("* preparing %s to accept the instance" % |
| 720 |
self.cfg.GetNodeName(self.target_node_uuid)) |
| 721 |
result = self.rpc.call_accept_instance(self.target_node_uuid, |
| 722 |
self.instance,
|
| 723 |
migration_info, |
| 724 |
self.nodes_ip[self.target_node_uuid]) |
| 725 |
|
| 726 |
msg = result.fail_msg |
| 727 |
if msg:
|
| 728 |
logging.error("Instance pre-migration failed, trying to revert"
|
| 729 |
" disk status: %s", msg)
|
| 730 |
self.feedback_fn("Pre-migration failed, aborting") |
| 731 |
self._AbortMigration()
|
| 732 |
self._RevertDiskStatus()
|
| 733 |
raise errors.OpExecError("Could not pre-migrate instance %s: %s" % |
| 734 |
(self.instance.name, msg))
|
| 735 |
|
| 736 |
self.feedback_fn("* migrating instance to %s" % |
| 737 |
self.cfg.GetNodeName(self.target_node_uuid)) |
| 738 |
cluster = self.cfg.GetClusterInfo()
|
| 739 |
result = self.rpc.call_instance_migrate(
|
| 740 |
self.source_node_uuid, cluster.cluster_name, self.instance, |
| 741 |
self.nodes_ip[self.target_node_uuid], self.live) |
| 742 |
msg = result.fail_msg |
| 743 |
if msg:
|
| 744 |
logging.error("Instance migration failed, trying to revert"
|
| 745 |
" disk status: %s", msg)
|
| 746 |
self.feedback_fn("Migration failed, aborting") |
| 747 |
self._AbortMigration()
|
| 748 |
self._RevertDiskStatus()
|
| 749 |
raise errors.OpExecError("Could not migrate instance %s: %s" % |
| 750 |
(self.instance.name, msg))
|
| 751 |
|
| 752 |
self.feedback_fn("* starting memory transfer") |
| 753 |
last_feedback = time.time() |
| 754 |
while True: |
| 755 |
result = self.rpc.call_instance_get_migration_status(
|
| 756 |
self.source_node_uuid, self.instance) |
| 757 |
msg = result.fail_msg |
| 758 |
ms = result.payload # MigrationStatus instance
|
| 759 |
if msg or (ms.status in constants.HV_MIGRATION_FAILED_STATUSES): |
| 760 |
logging.error("Instance migration failed, trying to revert"
|
| 761 |
" disk status: %s", msg)
|
| 762 |
self.feedback_fn("Migration failed, aborting") |
| 763 |
self._AbortMigration()
|
| 764 |
self._RevertDiskStatus()
|
| 765 |
if not msg: |
| 766 |
msg = "hypervisor returned failure"
|
| 767 |
raise errors.OpExecError("Could not migrate instance %s: %s" % |
| 768 |
(self.instance.name, msg))
|
| 769 |
|
| 770 |
if result.payload.status != constants.HV_MIGRATION_ACTIVE:
|
| 771 |
self.feedback_fn("* memory transfer complete") |
| 772 |
break
|
| 773 |
|
| 774 |
if (utils.TimeoutExpired(last_feedback,
|
| 775 |
self._MIGRATION_FEEDBACK_INTERVAL) and |
| 776 |
ms.transferred_ram is not None): |
| 777 |
mem_progress = 100 * float(ms.transferred_ram) / float(ms.total_ram) |
| 778 |
self.feedback_fn("* memory transfer progress: %.2f %%" % mem_progress) |
| 779 |
last_feedback = time.time() |
| 780 |
|
| 781 |
time.sleep(self._MIGRATION_POLL_INTERVAL)
|
| 782 |
|
| 783 |
result = self.rpc.call_instance_finalize_migration_src(
|
| 784 |
self.source_node_uuid, self.instance, True, self.live) |
| 785 |
msg = result.fail_msg |
| 786 |
if msg:
|
| 787 |
logging.error("Instance migration succeeded, but finalization failed"
|
| 788 |
" on the source node: %s", msg)
|
| 789 |
raise errors.OpExecError("Could not finalize instance migration: %s" % |
| 790 |
msg) |
| 791 |
|
| 792 |
self.instance.primary_node = self.target_node_uuid |
| 793 |
|
| 794 |
# distribute new instance config to the other nodes
|
| 795 |
self.cfg.Update(self.instance, self.feedback_fn) |
| 796 |
|
| 797 |
result = self.rpc.call_instance_finalize_migration_dst(
|
| 798 |
self.target_node_uuid, self.instance, migration_info, True) |
| 799 |
msg = result.fail_msg |
| 800 |
if msg:
|
| 801 |
logging.error("Instance migration succeeded, but finalization failed"
|
| 802 |
" on the target node: %s", msg)
|
| 803 |
raise errors.OpExecError("Could not finalize instance migration: %s" % |
| 804 |
msg) |
| 805 |
|
| 806 |
if self.instance.disk_template not in constants.DTS_EXT_MIRROR: |
| 807 |
self._EnsureSecondary(self.source_node_uuid) |
| 808 |
self._WaitUntilSync()
|
| 809 |
self._GoStandalone()
|
| 810 |
self._GoReconnect(False) |
| 811 |
self._WaitUntilSync()
|
| 812 |
|
| 813 |
# If the instance's disk template is `rbd' or `ext' and there was a
|
| 814 |
# successful migration, unmap the device from the source node.
|
| 815 |
if self.instance.disk_template in (constants.DT_RBD, constants.DT_EXT): |
| 816 |
disks = ExpandCheckDisks(self.instance, self.instance.disks) |
| 817 |
self.feedback_fn("* unmapping instance's disks from %s" % |
| 818 |
self.cfg.GetNodeName(self.source_node_uuid)) |
| 819 |
for disk in disks: |
| 820 |
result = self.rpc.call_blockdev_shutdown(self.source_node_uuid, |
| 821 |
(disk, self.instance))
|
| 822 |
msg = result.fail_msg |
| 823 |
if msg:
|
| 824 |
logging.error("Migration was successful, but couldn't unmap the"
|
| 825 |
" block device %s on source node %s: %s",
|
| 826 |
disk.iv_name, |
| 827 |
self.cfg.GetNodeName(self.source_node_uuid), msg) |
| 828 |
logging.error("You need to unmap the device %s manually on %s",
|
| 829 |
disk.iv_name, |
| 830 |
self.cfg.GetNodeName(self.source_node_uuid)) |
| 831 |
|
| 832 |
self.feedback_fn("* done") |
| 833 |
|
| 834 |
def _ExecFailover(self): |
| 835 |
"""Failover an instance.
|
| 836 |
|
| 837 |
The failover is done by shutting it down on its present node and
|
| 838 |
starting it on the secondary.
|
| 839 |
|
| 840 |
"""
|
| 841 |
primary_node = self.cfg.GetNodeInfo(self.instance.primary_node) |
| 842 |
|
| 843 |
source_node_uuid = self.instance.primary_node
|
| 844 |
|
| 845 |
if self.instance.disks_active: |
| 846 |
self.feedback_fn("* checking disk consistency between source and target") |
| 847 |
for (idx, dev) in enumerate(self.instance.disks): |
| 848 |
# for drbd, these are drbd over lvm
|
| 849 |
if not CheckDiskConsistency(self.lu, self.instance, dev, |
| 850 |
self.target_node_uuid, False): |
| 851 |
if primary_node.offline:
|
| 852 |
self.feedback_fn("Node %s is offline, ignoring degraded disk %s on" |
| 853 |
" target node %s" %
|
| 854 |
(primary_node.name, idx, |
| 855 |
self.cfg.GetNodeName(self.target_node_uuid))) |
| 856 |
elif not self.ignore_consistency: |
| 857 |
raise errors.OpExecError("Disk %s is degraded on target node," |
| 858 |
" aborting failover" % idx)
|
| 859 |
else:
|
| 860 |
self.feedback_fn("* not checking disk consistency as instance is not" |
| 861 |
" running")
|
| 862 |
|
| 863 |
self.feedback_fn("* shutting down instance on source node") |
| 864 |
logging.info("Shutting down instance %s on node %s",
|
| 865 |
self.instance.name, self.cfg.GetNodeName(source_node_uuid)) |
| 866 |
|
| 867 |
result = self.rpc.call_instance_shutdown(source_node_uuid, self.instance, |
| 868 |
self.shutdown_timeout,
|
| 869 |
self.lu.op.reason)
|
| 870 |
msg = result.fail_msg |
| 871 |
if msg:
|
| 872 |
if self.ignore_consistency or primary_node.offline: |
| 873 |
self.lu.LogWarning("Could not shutdown instance %s on node %s," |
| 874 |
" proceeding anyway; please make sure node"
|
| 875 |
" %s is down; error details: %s",
|
| 876 |
self.instance.name,
|
| 877 |
self.cfg.GetNodeName(source_node_uuid),
|
| 878 |
self.cfg.GetNodeName(source_node_uuid), msg)
|
| 879 |
else:
|
| 880 |
raise errors.OpExecError("Could not shutdown instance %s on" |
| 881 |
" node %s: %s" %
|
| 882 |
(self.instance.name,
|
| 883 |
self.cfg.GetNodeName(source_node_uuid), msg))
|
| 884 |
|
| 885 |
self.feedback_fn("* deactivating the instance's disks on source node") |
| 886 |
if not ShutdownInstanceDisks(self.lu, self.instance, ignore_primary=True): |
| 887 |
raise errors.OpExecError("Can't shut down the instance's disks") |
| 888 |
|
| 889 |
self.instance.primary_node = self.target_node_uuid |
| 890 |
# distribute new instance config to the other nodes
|
| 891 |
self.cfg.Update(self.instance, self.feedback_fn) |
| 892 |
|
| 893 |
# Only start the instance if it's marked as up
|
| 894 |
if self.instance.admin_state == constants.ADMINST_UP: |
| 895 |
self.feedback_fn("* activating the instance's disks on target node %s" % |
| 896 |
self.cfg.GetNodeName(self.target_node_uuid)) |
| 897 |
logging.info("Starting instance %s on node %s", self.instance.name, |
| 898 |
self.cfg.GetNodeName(self.target_node_uuid)) |
| 899 |
|
| 900 |
disks_ok, _ = AssembleInstanceDisks(self.lu, self.instance, |
| 901 |
ignore_secondaries=True)
|
| 902 |
if not disks_ok: |
| 903 |
ShutdownInstanceDisks(self.lu, self.instance) |
| 904 |
raise errors.OpExecError("Can't activate the instance's disks") |
| 905 |
|
| 906 |
self.feedback_fn("* starting the instance on the target node %s" % |
| 907 |
self.cfg.GetNodeName(self.target_node_uuid)) |
| 908 |
result = self.rpc.call_instance_start(self.target_node_uuid, |
| 909 |
(self.instance, None, None), False, |
| 910 |
self.lu.op.reason)
|
| 911 |
msg = result.fail_msg |
| 912 |
if msg:
|
| 913 |
ShutdownInstanceDisks(self.lu, self.instance) |
| 914 |
raise errors.OpExecError("Could not start instance %s on node %s: %s" % |
| 915 |
(self.instance.name,
|
| 916 |
self.cfg.GetNodeName(self.target_node_uuid), |
| 917 |
msg)) |
| 918 |
|
| 919 |
def Exec(self, feedback_fn): |
| 920 |
"""Perform the migration.
|
| 921 |
|
| 922 |
"""
|
| 923 |
self.feedback_fn = feedback_fn
|
| 924 |
self.source_node_uuid = self.instance.primary_node |
| 925 |
|
| 926 |
# FIXME: if we implement migrate-to-any in DRBD, this needs fixing
|
| 927 |
if self.instance.disk_template in constants.DTS_INT_MIRROR: |
| 928 |
self.target_node_uuid = self.instance.secondary_nodes[0] |
| 929 |
# Otherwise self.target_node has been populated either
|
| 930 |
# directly, or through an iallocator.
|
| 931 |
|
| 932 |
self.all_node_uuids = [self.source_node_uuid, self.target_node_uuid] |
| 933 |
self.nodes_ip = dict((uuid, node.secondary_ip) for (uuid, node) |
| 934 |
in self.cfg.GetMultiNodeInfo(self.all_node_uuids)) |
| 935 |
|
| 936 |
if self.failover: |
| 937 |
feedback_fn("Failover instance %s" % self.instance.name) |
| 938 |
self._ExecFailover()
|
| 939 |
else:
|
| 940 |
feedback_fn("Migrating instance %s" % self.instance.name) |
| 941 |
|
| 942 |
if self.cleanup: |
| 943 |
return self._ExecCleanup() |
| 944 |
else:
|
| 945 |
return self._ExecMigration() |