4 # Copyright (C) 2006, 2007 Google Inc.
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.
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.
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
22 """Module implementing the logic behind the cluster operations
24 This module implements the logic for doing operations in the cluster. There
25 are two kinds of classes defined:
26 - logical units, which know how to deal with their specific opcode only
27 - the processor, which dispatches the opcodes to their logical units
35 from ganeti import opcodes
36 from ganeti import constants
37 from ganeti import errors
38 from ganeti import rpc
39 from ganeti import cmdlib
40 from ganeti import locking
41 from ganeti import utils
44 class LockAcquireTimeout(Exception):
45 """Exception to report timeouts on acquiring locks.
50 def _CalculateLockAttemptTimeouts():
51 """Calculate timeouts for lock attempts.
56 # Wait for a total of at least 150s before doing a blocking acquire
57 while sum(result) < 150.0:
58 timeout = (result[-1] * 1.05) ** 1.25
60 # Cap timeout at 10 seconds. This gives other jobs a chance to run
61 # even if we're still trying to get our locks, before finally moving
62 # to a blocking acquire.
67 # Lower boundary for safety
70 result.append(timeout)
75 class LockAttemptTimeoutStrategy(object):
76 """Class with lock acquire timeout strategy.
85 _TIMEOUT_PER_ATTEMPT = _CalculateLockAttemptTimeouts()
87 def __init__(self, _time_fn=time.time, _random_fn=random.random):
88 """Initializes this class.
90 @param _time_fn: Time function for unittests
91 @param _random_fn: Random number generator for unittests
96 self._timeouts = iter(self._TIMEOUT_PER_ATTEMPT)
97 self._time_fn = _time_fn
98 self._random_fn = _random_fn
100 def NextAttempt(self):
101 """Returns the timeout for the next attempt.
105 timeout = self._timeouts.next()
106 except StopIteration:
107 # No more timeouts, do blocking acquire
110 if timeout is not None:
111 # Add a small variation (-/+ 5%) to timeout. This helps in situations
112 # where two or more jobs are fighting for the same lock(s).
113 variation_range = timeout * 0.1
114 timeout += ((self._random_fn() * variation_range) -
115 (variation_range * 0.5))
120 class OpExecCbBase: # pylint: disable-msg=W0232
121 """Base class for OpCode execution callbacks.
124 def NotifyStart(self):
125 """Called when we are about to execute the LU.
127 This function is called when we're about to start the lu's Exec() method,
128 that is, after we have acquired all locks.
132 def Feedback(self, *args):
133 """Sends feedback from the LU code to the end-user.
137 def CheckCancel(self):
138 """Check whether job has been cancelled.
143 class Processor(object):
144 """Object which runs OpCodes"""
147 opcodes.OpPostInitCluster: cmdlib.LUPostInitCluster,
148 opcodes.OpDestroyCluster: cmdlib.LUDestroyCluster,
149 opcodes.OpQueryClusterInfo: cmdlib.LUQueryClusterInfo,
150 opcodes.OpVerifyCluster: cmdlib.LUVerifyCluster,
151 opcodes.OpQueryConfigValues: cmdlib.LUQueryConfigValues,
152 opcodes.OpRenameCluster: cmdlib.LURenameCluster,
153 opcodes.OpVerifyDisks: cmdlib.LUVerifyDisks,
154 opcodes.OpSetClusterParams: cmdlib.LUSetClusterParams,
155 opcodes.OpRedistributeConfig: cmdlib.LURedistributeConfig,
156 opcodes.OpRepairDiskSizes: cmdlib.LURepairDiskSizes,
157 opcodes.OpQuery: cmdlib.LUQuery,
158 opcodes.OpQueryFields: cmdlib.LUQueryFields,
160 opcodes.OpAddNode: cmdlib.LUAddNode,
161 opcodes.OpQueryNodes: cmdlib.LUQueryNodes,
162 opcodes.OpQueryNodeVolumes: cmdlib.LUQueryNodeVolumes,
163 opcodes.OpQueryNodeStorage: cmdlib.LUQueryNodeStorage,
164 opcodes.OpModifyNodeStorage: cmdlib.LUModifyNodeStorage,
165 opcodes.OpRepairNodeStorage: cmdlib.LURepairNodeStorage,
166 opcodes.OpRemoveNode: cmdlib.LURemoveNode,
167 opcodes.OpSetNodeParams: cmdlib.LUSetNodeParams,
168 opcodes.OpPowercycleNode: cmdlib.LUPowercycleNode,
169 opcodes.OpMigrateNode: cmdlib.LUMigrateNode,
170 opcodes.OpNodeEvacuationStrategy: cmdlib.LUNodeEvacuationStrategy,
172 opcodes.OpCreateInstance: cmdlib.LUCreateInstance,
173 opcodes.OpReinstallInstance: cmdlib.LUReinstallInstance,
174 opcodes.OpRemoveInstance: cmdlib.LURemoveInstance,
175 opcodes.OpRenameInstance: cmdlib.LURenameInstance,
176 opcodes.OpActivateInstanceDisks: cmdlib.LUActivateInstanceDisks,
177 opcodes.OpShutdownInstance: cmdlib.LUShutdownInstance,
178 opcodes.OpStartupInstance: cmdlib.LUStartupInstance,
179 opcodes.OpRebootInstance: cmdlib.LURebootInstance,
180 opcodes.OpDeactivateInstanceDisks: cmdlib.LUDeactivateInstanceDisks,
181 opcodes.OpReplaceDisks: cmdlib.LUReplaceDisks,
182 opcodes.OpRecreateInstanceDisks: cmdlib.LURecreateInstanceDisks,
183 opcodes.OpFailoverInstance: cmdlib.LUFailoverInstance,
184 opcodes.OpMigrateInstance: cmdlib.LUMigrateInstance,
185 opcodes.OpMoveInstance: cmdlib.LUMoveInstance,
186 opcodes.OpConnectConsole: cmdlib.LUConnectConsole,
187 opcodes.OpQueryInstances: cmdlib.LUQueryInstances,
188 opcodes.OpQueryInstanceData: cmdlib.LUQueryInstanceData,
189 opcodes.OpSetInstanceParams: cmdlib.LUSetInstanceParams,
190 opcodes.OpGrowDisk: cmdlib.LUGrowDisk,
192 opcodes.OpQueryGroups: cmdlib.LUQueryGroups,
194 opcodes.OpDiagnoseOS: cmdlib.LUDiagnoseOS,
196 opcodes.OpQueryExports: cmdlib.LUQueryExports,
197 opcodes.OpPrepareExport: cmdlib.LUPrepareExport,
198 opcodes.OpExportInstance: cmdlib.LUExportInstance,
199 opcodes.OpRemoveExport: cmdlib.LURemoveExport,
201 opcodes.OpGetTags: cmdlib.LUGetTags,
202 opcodes.OpSearchTags: cmdlib.LUSearchTags,
203 opcodes.OpAddTags: cmdlib.LUAddTags,
204 opcodes.OpDelTags: cmdlib.LUDelTags,
206 opcodes.OpTestDelay: cmdlib.LUTestDelay,
207 opcodes.OpTestAllocator: cmdlib.LUTestAllocator,
208 opcodes.OpTestJobqueue: cmdlib.LUTestJobqueue,
210 opcodes.OpOutOfBand: cmdlib.LUOutOfBand,
213 def __init__(self, context, ec_id):
214 """Constructor for Processor
216 @type context: GanetiContext
217 @param context: global Ganeti context
219 @param ec_id: execution context identifier
222 self.context = context
225 self.rpc = rpc.RpcRunner(context.cfg)
226 self.hmclass = HooksMaster
228 def _AcquireLocks(self, level, names, shared, timeout, priority):
229 """Acquires locks via the Ganeti lock manager.
232 @param level: Lock level
233 @type names: list or string
234 @param names: Lock names
236 @param shared: Whether the locks should be acquired in shared mode
237 @type timeout: None or float
238 @param timeout: Timeout for acquiring the locks
239 @raise LockAcquireTimeout: In case locks couldn't be acquired in specified
244 self._cbs.CheckCancel()
246 acquired = self.context.glm.acquire(level, names, shared=shared,
247 timeout=timeout, priority=priority)
250 raise LockAcquireTimeout()
254 def _ExecLU(self, lu):
255 """Logical Unit execution sequence.
258 write_count = self.context.cfg.write_count
260 hm = HooksMaster(self.rpc.call_hooks_runner, lu)
261 h_results = hm.RunPhase(constants.HOOKS_PHASE_PRE)
262 lu.HooksCallBack(constants.HOOKS_PHASE_PRE, h_results,
265 if getattr(lu.op, "dry_run", False):
266 # in this mode, no post-hooks are run, and the config is not
267 # written (as it might have been modified by another LU, and we
268 # shouldn't do writeout on behalf of other threads
269 self.LogInfo("dry-run mode requested, not actually executing"
271 return lu.dry_run_result
274 result = lu.Exec(self.Log)
275 h_results = hm.RunPhase(constants.HOOKS_PHASE_POST)
276 result = lu.HooksCallBack(constants.HOOKS_PHASE_POST, h_results,
279 # FIXME: This needs locks if not lu_class.REQ_BGL
280 if write_count != self.context.cfg.write_count:
285 def _LockAndExecLU(self, lu, level, calc_timeout, priority):
286 """Execute a Logical Unit, with the needed locks.
288 This is a recursive function that starts locking the given level, and
289 proceeds up, till there are no more locks to acquire. Then it executes the
290 given LU and its opcodes.
293 adding_locks = level in lu.add_locks
294 acquiring_locks = level in lu.needed_locks
295 if level not in locking.LEVELS:
297 self._cbs.NotifyStart()
299 result = self._ExecLU(lu)
301 elif adding_locks and acquiring_locks:
302 # We could both acquire and add locks at the same level, but for now we
303 # don't need this, so we'll avoid the complicated code needed.
304 raise NotImplementedError("Can't declare locks to acquire when adding"
307 elif adding_locks or acquiring_locks:
308 lu.DeclareLocks(level)
309 share = lu.share_locks[level]
312 assert adding_locks ^ acquiring_locks, \
313 "Locks must be either added or acquired"
317 needed_locks = lu.needed_locks[level]
319 acquired = self._AcquireLocks(level, needed_locks, share,
320 calc_timeout(), priority)
323 add_locks = lu.add_locks[level]
324 lu.remove_locks[level] = add_locks
327 self.context.glm.add(level, add_locks, acquired=1, shared=share)
328 except errors.LockError:
329 raise errors.OpPrereqError(
330 "Couldn't add locks (%s), probably because of a race condition"
331 " with another job, who added them first" % add_locks,
337 lu.acquired_locks[level] = acquired
339 result = self._LockAndExecLU(lu, level + 1, calc_timeout, priority)
341 if level in lu.remove_locks:
342 self.context.glm.remove(level, lu.remove_locks[level])
344 if self.context.glm.is_owned(level):
345 self.context.glm.release(level)
348 result = self._LockAndExecLU(lu, level + 1, calc_timeout, priority)
352 def ExecOpCode(self, op, cbs, timeout=None, priority=None):
353 """Execute an opcode.
355 @type op: an OpCode instance
356 @param op: the opcode to be executed
357 @type cbs: L{OpExecCbBase}
358 @param cbs: Runtime callbacks
359 @type timeout: float or None
360 @param timeout: Maximum time to acquire all locks, None for no timeout
361 @type priority: number or None
362 @param priority: Priority for acquiring lock(s)
363 @raise LockAcquireTimeout: In case locks couldn't be acquired in specified
367 if not isinstance(op, opcodes.OpCode):
368 raise errors.ProgrammerError("Non-opcode instance passed"
371 lu_class = self.DISPATCH_TABLE.get(op.__class__, None)
373 raise errors.OpCodeUnknown("Unknown opcode")
376 calc_timeout = lambda: None
378 calc_timeout = utils.RunningTimeout(timeout, False).Remaining
382 # Acquire the Big Ganeti Lock exclusively if this LU requires it,
383 # and in a shared fashion otherwise (to prevent concurrent run with
385 self._AcquireLocks(locking.LEVEL_CLUSTER, locking.BGL,
386 not lu_class.REQ_BGL, calc_timeout(),
389 lu = lu_class(self, op, self.context, self.rpc)
391 assert lu.needed_locks is not None, "needed_locks not set by LU"
394 return self._LockAndExecLU(lu, locking.LEVEL_INSTANCE, calc_timeout,
398 self.context.cfg.DropECReservations(self._ec_id)
400 self.context.glm.release(locking.LEVEL_CLUSTER)
404 def Log(self, *args):
405 """Forward call to feedback callback function.
409 self._cbs.Feedback(*args)
411 def LogStep(self, current, total, message):
412 """Log a change in LU execution progress.
415 logging.debug("Step %d/%d %s", current, total, message)
416 self.Log("STEP %d/%d %s" % (current, total, message))
418 def LogWarning(self, message, *args, **kwargs):
419 """Log a warning to the logs and the user.
421 The optional keyword argument is 'hint' and can be used to show a
422 hint to the user (presumably related to the warning). If the
423 message is empty, it will not be printed at all, allowing one to
427 assert not kwargs or (len(kwargs) == 1 and "hint" in kwargs), \
428 "Invalid keyword arguments for LogWarning (%s)" % str(kwargs)
430 message = message % tuple(args)
432 logging.warning(message)
433 self.Log(" - WARNING: %s" % message)
435 self.Log(" Hint: %s" % kwargs["hint"])
437 def LogInfo(self, message, *args):
438 """Log an informational message to the logs and the user.
442 message = message % tuple(args)
443 logging.info(message)
444 self.Log(" - INFO: %s" % message)
448 errors.ProgrammerError("Tried to use execution context id when not set")
452 class HooksMaster(object):
455 This class distributes the run commands to the nodes based on the
458 In order to remove the direct dependency on the rpc module, the
459 constructor needs a function which actually does the remote
460 call. This will usually be rpc.call_hooks_runner, but any function
461 which behaves the same works.
464 def __init__(self, callfn, lu):
468 self.env, node_list_pre, node_list_post = self._BuildEnv()
469 self.node_list = {constants.HOOKS_PHASE_PRE: node_list_pre,
470 constants.HOOKS_PHASE_POST: node_list_post}
473 """Compute the environment and the target nodes.
475 Based on the opcode and the current node list, this builds the
476 environment for the hooks and the target node list for the run.
480 "PATH": "/sbin:/bin:/usr/sbin:/usr/bin",
481 "GANETI_HOOKS_VERSION": constants.HOOKS_VERSION,
482 "GANETI_OP_CODE": self.op.OP_ID,
483 "GANETI_OBJECT_TYPE": self.lu.HTYPE,
484 "GANETI_DATA_DIR": constants.DATA_DIR,
487 if self.lu.HPATH is not None:
488 lu_env, lu_nodes_pre, lu_nodes_post = self.lu.BuildHooksEnv()
491 env["GANETI_" + key] = lu_env[key]
493 lu_nodes_pre = lu_nodes_post = []
495 return env, frozenset(lu_nodes_pre), frozenset(lu_nodes_post)
497 def _RunWrapper(self, node_list, hpath, phase):
498 """Simple wrapper over self.callfn.
500 This method fixes the environment before doing the rpc call.
503 env = self.env.copy()
504 env["GANETI_HOOKS_PHASE"] = phase
505 env["GANETI_HOOKS_PATH"] = hpath
506 if self.lu.cfg is not None:
507 env["GANETI_CLUSTER"] = self.lu.cfg.GetClusterName()
508 env["GANETI_MASTER"] = self.lu.cfg.GetMasterNode()
510 env = dict([(str(key), str(val)) for key, val in env.iteritems()])
512 return self.callfn(node_list, hpath, phase, env)
514 def RunPhase(self, phase, nodes=None):
515 """Run all the scripts for a phase.
517 This is the main function of the HookMaster.
519 @param phase: one of L{constants.HOOKS_PHASE_POST} or
520 L{constants.HOOKS_PHASE_PRE}; it denotes the hooks phase
521 @param nodes: overrides the predefined list of nodes for the given phase
522 @return: the processed results of the hooks multi-node rpc call
523 @raise errors.HooksFailure: on communication failure to the nodes
524 @raise errors.HooksAbort: on failure of one of the hooks
527 if not self.node_list[phase] and not nodes:
528 # empty node list, we should not attempt to run this as either
529 # we're in the cluster init phase and the rpc client part can't
530 # even attempt to run, or this LU doesn't do hooks at all
532 hpath = self.lu.HPATH
533 if nodes is not None:
534 results = self._RunWrapper(nodes, hpath, phase)
536 results = self._RunWrapper(self.node_list[phase], hpath, phase)
539 msg = "Communication Failure"
540 if phase == constants.HOOKS_PHASE_PRE:
541 raise errors.HooksFailure(msg)
543 self.lu.LogWarning(msg)
545 for node_name in results:
546 res = results[node_name]
551 self.lu.LogWarning("Communication failure to node %s: %s",
554 for script, hkr, output in res.payload:
555 if hkr == constants.HKR_FAIL:
556 if phase == constants.HOOKS_PHASE_PRE:
557 errs.append((node_name, script, output))
560 output = "(no output)"
561 self.lu.LogWarning("On %s script %s failed, output: %s" %
562 (node_name, script, output))
563 if errs and phase == constants.HOOKS_PHASE_PRE:
564 raise errors.HooksAbort(errs)
567 def RunConfigUpdate(self):
568 """Run the special configuration update hook
570 This is a special hook that runs only on the master after each
571 top-level LI if the configuration has been updated.
574 phase = constants.HOOKS_PHASE_POST
575 hpath = constants.HOOKS_NAME_CFGUPDATE
576 nodes = [self.lu.cfg.GetMasterNode()]
577 self._RunWrapper(nodes, hpath, phase)