root / lib / watcher / __init__.py @ c300dbe4
History | View | Annotate | Download (24.3 kB)
1 |
#
|
---|---|
2 |
#
|
3 |
|
4 |
# Copyright (C) 2006, 2007, 2008, 2009, 2010, 2011, 2012 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 |
"""Tool to restart erroneously downed virtual machines.
|
23 |
|
24 |
This program and set of classes implement a watchdog to restart
|
25 |
virtual machines in a Ganeti cluster that have crashed or been killed
|
26 |
by a node reboot. Run from cron or similar.
|
27 |
|
28 |
"""
|
29 |
|
30 |
import os |
31 |
import os.path |
32 |
import sys |
33 |
import time |
34 |
import logging |
35 |
import operator |
36 |
import errno |
37 |
from optparse import OptionParser |
38 |
|
39 |
from ganeti import utils |
40 |
from ganeti import constants |
41 |
from ganeti import compat |
42 |
from ganeti import errors |
43 |
from ganeti import opcodes |
44 |
from ganeti import cli |
45 |
from ganeti import luxi |
46 |
from ganeti import rapi |
47 |
from ganeti import netutils |
48 |
from ganeti import qlang |
49 |
from ganeti import objects |
50 |
from ganeti import ssconf |
51 |
from ganeti import ht |
52 |
from ganeti import pathutils |
53 |
|
54 |
import ganeti.rapi.client # pylint: disable=W0611 |
55 |
from ganeti.rapi.client import UsesRapiClient |
56 |
|
57 |
from ganeti.watcher import nodemaint |
58 |
from ganeti.watcher import state |
59 |
|
60 |
|
61 |
MAXTRIES = 5
|
62 |
BAD_STATES = compat.UniqueFrozenset([ |
63 |
constants.INSTST_ERRORDOWN, |
64 |
]) |
65 |
HELPLESS_STATES = compat.UniqueFrozenset([ |
66 |
constants.INSTST_NODEDOWN, |
67 |
constants.INSTST_NODEOFFLINE, |
68 |
]) |
69 |
NOTICE = "NOTICE"
|
70 |
ERROR = "ERROR"
|
71 |
|
72 |
#: Number of seconds to wait between starting child processes for node groups
|
73 |
CHILD_PROCESS_DELAY = 1.0
|
74 |
|
75 |
#: How many seconds to wait for instance status file lock
|
76 |
INSTANCE_STATUS_LOCK_TIMEOUT = 10.0
|
77 |
|
78 |
|
79 |
class NotMasterError(errors.GenericError): |
80 |
"""Exception raised when this host is not the master."""
|
81 |
|
82 |
|
83 |
def ShouldPause(): |
84 |
"""Check whether we should pause.
|
85 |
|
86 |
"""
|
87 |
return bool(utils.ReadWatcherPauseFile(pathutils.WATCHER_PAUSEFILE)) |
88 |
|
89 |
|
90 |
def StartNodeDaemons(): |
91 |
"""Start all the daemons that should be running on all nodes.
|
92 |
|
93 |
"""
|
94 |
# on master or not, try to start the node daemon
|
95 |
utils.EnsureDaemon(constants.NODED) |
96 |
# start confd as well. On non candidates it will be in disabled mode.
|
97 |
if constants.ENABLE_CONFD:
|
98 |
utils.EnsureDaemon(constants.CONFD) |
99 |
# start mond as well: all nodes need monitoring
|
100 |
if constants.ENABLE_MOND:
|
101 |
utils.EnsureDaemon(constants.MOND) |
102 |
|
103 |
|
104 |
|
105 |
def RunWatcherHooks(): |
106 |
"""Run the watcher hooks.
|
107 |
|
108 |
"""
|
109 |
hooks_dir = utils.PathJoin(pathutils.HOOKS_BASE_DIR, |
110 |
constants.HOOKS_NAME_WATCHER) |
111 |
if not os.path.isdir(hooks_dir): |
112 |
return
|
113 |
|
114 |
try:
|
115 |
results = utils.RunParts(hooks_dir) |
116 |
except Exception, err: # pylint: disable=W0703 |
117 |
logging.exception("RunParts %s failed: %s", hooks_dir, err)
|
118 |
return
|
119 |
|
120 |
for (relname, status, runresult) in results: |
121 |
if status == constants.RUNPARTS_SKIP:
|
122 |
logging.debug("Watcher hook %s: skipped", relname)
|
123 |
elif status == constants.RUNPARTS_ERR:
|
124 |
logging.warning("Watcher hook %s: error (%s)", relname, runresult)
|
125 |
elif status == constants.RUNPARTS_RUN:
|
126 |
if runresult.failed:
|
127 |
logging.warning("Watcher hook %s: failed (exit: %d) (output: %s)",
|
128 |
relname, runresult.exit_code, runresult.output) |
129 |
else:
|
130 |
logging.debug("Watcher hook %s: success (output: %s)", relname,
|
131 |
runresult.output) |
132 |
else:
|
133 |
raise errors.ProgrammerError("Unknown status %s returned by RunParts", |
134 |
status) |
135 |
|
136 |
|
137 |
class Instance(object): |
138 |
"""Abstraction for a Virtual Machine instance.
|
139 |
|
140 |
"""
|
141 |
def __init__(self, name, status, autostart, snodes): |
142 |
self.name = name
|
143 |
self.status = status
|
144 |
self.autostart = autostart
|
145 |
self.snodes = snodes
|
146 |
|
147 |
def Restart(self, cl): |
148 |
"""Encapsulates the start of an instance.
|
149 |
|
150 |
"""
|
151 |
op = opcodes.OpInstanceStartup(instance_name=self.name, force=False) |
152 |
cli.SubmitOpCode(op, cl=cl) |
153 |
|
154 |
def ActivateDisks(self, cl): |
155 |
"""Encapsulates the activation of all disks of an instance.
|
156 |
|
157 |
"""
|
158 |
op = opcodes.OpInstanceActivateDisks(instance_name=self.name)
|
159 |
cli.SubmitOpCode(op, cl=cl) |
160 |
|
161 |
|
162 |
class Node: |
163 |
"""Data container representing cluster node.
|
164 |
|
165 |
"""
|
166 |
def __init__(self, name, bootid, offline, secondaries): |
167 |
"""Initializes this class.
|
168 |
|
169 |
"""
|
170 |
self.name = name
|
171 |
self.bootid = bootid
|
172 |
self.offline = offline
|
173 |
self.secondaries = secondaries
|
174 |
|
175 |
|
176 |
def _CheckInstances(cl, notepad, instances): |
177 |
"""Make a pass over the list of instances, restarting downed ones.
|
178 |
|
179 |
"""
|
180 |
notepad.MaintainInstanceList(instances.keys()) |
181 |
|
182 |
started = set()
|
183 |
|
184 |
for inst in instances.values(): |
185 |
if inst.status in BAD_STATES: |
186 |
n = notepad.NumberOfRestartAttempts(inst.name) |
187 |
|
188 |
if n > MAXTRIES:
|
189 |
logging.warning("Not restarting instance '%s', retries exhausted",
|
190 |
inst.name) |
191 |
continue
|
192 |
|
193 |
if n == MAXTRIES:
|
194 |
notepad.RecordRestartAttempt(inst.name) |
195 |
logging.error("Could not restart instance '%s' after %s attempts,"
|
196 |
" giving up", inst.name, MAXTRIES)
|
197 |
continue
|
198 |
|
199 |
try:
|
200 |
logging.info("Restarting instance '%s' (attempt #%s)",
|
201 |
inst.name, n + 1)
|
202 |
inst.Restart(cl) |
203 |
except Exception: # pylint: disable=W0703 |
204 |
logging.exception("Error while restarting instance '%s'", inst.name)
|
205 |
else:
|
206 |
started.add(inst.name) |
207 |
|
208 |
notepad.RecordRestartAttempt(inst.name) |
209 |
|
210 |
else:
|
211 |
if notepad.NumberOfRestartAttempts(inst.name):
|
212 |
notepad.RemoveInstance(inst.name) |
213 |
if inst.status not in HELPLESS_STATES: |
214 |
logging.info("Restart of instance '%s' succeeded", inst.name)
|
215 |
|
216 |
return started
|
217 |
|
218 |
|
219 |
def _CheckDisks(cl, notepad, nodes, instances, started): |
220 |
"""Check all nodes for restarted ones.
|
221 |
|
222 |
"""
|
223 |
check_nodes = [] |
224 |
|
225 |
for node in nodes.values(): |
226 |
old = notepad.GetNodeBootID(node.name) |
227 |
if not node.bootid: |
228 |
# Bad node, not returning a boot id
|
229 |
if not node.offline: |
230 |
logging.debug("Node '%s' missing boot ID, skipping secondary checks",
|
231 |
node.name) |
232 |
continue
|
233 |
|
234 |
if old != node.bootid:
|
235 |
# Node's boot ID has changed, probably through a reboot
|
236 |
check_nodes.append(node) |
237 |
|
238 |
if check_nodes:
|
239 |
# Activate disks for all instances with any of the checked nodes as a
|
240 |
# secondary node.
|
241 |
for node in check_nodes: |
242 |
for instance_name in node.secondaries: |
243 |
try:
|
244 |
inst = instances[instance_name] |
245 |
except KeyError: |
246 |
logging.info("Can't find instance '%s', maybe it was ignored",
|
247 |
instance_name) |
248 |
continue
|
249 |
|
250 |
if not inst.autostart: |
251 |
logging.info("Skipping disk activation for non-autostart"
|
252 |
" instance '%s'", inst.name)
|
253 |
continue
|
254 |
|
255 |
if inst.name in started: |
256 |
# we already tried to start the instance, which should have
|
257 |
# activated its drives (if they can be at all)
|
258 |
logging.debug("Skipping disk activation for instance '%s' as"
|
259 |
" it was already started", inst.name)
|
260 |
continue
|
261 |
|
262 |
try:
|
263 |
logging.info("Activating disks for instance '%s'", inst.name)
|
264 |
inst.ActivateDisks(cl) |
265 |
except Exception: # pylint: disable=W0703 |
266 |
logging.exception("Error while activating disks for instance '%s'",
|
267 |
inst.name) |
268 |
|
269 |
# Keep changed boot IDs
|
270 |
for node in check_nodes: |
271 |
notepad.SetNodeBootID(node.name, node.bootid) |
272 |
|
273 |
|
274 |
def _CheckForOfflineNodes(nodes, instance): |
275 |
"""Checks if given instances has any secondary in offline status.
|
276 |
|
277 |
@param instance: The instance object
|
278 |
@return: True if any of the secondary is offline, False otherwise
|
279 |
|
280 |
"""
|
281 |
return compat.any(nodes[node_name].offline for node_name in instance.snodes) |
282 |
|
283 |
|
284 |
def _VerifyDisks(cl, uuid, nodes, instances): |
285 |
"""Run a per-group "gnt-cluster verify-disks".
|
286 |
|
287 |
"""
|
288 |
job_id = cl.SubmitJob([opcodes.OpGroupVerifyDisks(group_name=uuid)]) |
289 |
((_, offline_disk_instances, _), ) = \ |
290 |
cli.PollJob(job_id, cl=cl, feedback_fn=logging.debug) |
291 |
cl.ArchiveJob(job_id) |
292 |
|
293 |
if not offline_disk_instances: |
294 |
# nothing to do
|
295 |
logging.debug("Verify-disks reported no offline disks, nothing to do")
|
296 |
return
|
297 |
|
298 |
logging.debug("Will activate disks for instance(s) %s",
|
299 |
utils.CommaJoin(offline_disk_instances)) |
300 |
|
301 |
# We submit only one job, and wait for it. Not optimal, but this puts less
|
302 |
# load on the job queue.
|
303 |
job = [] |
304 |
for name in offline_disk_instances: |
305 |
try:
|
306 |
inst = instances[name] |
307 |
except KeyError: |
308 |
logging.info("Can't find instance '%s', maybe it was ignored", name)
|
309 |
continue
|
310 |
|
311 |
if inst.status in HELPLESS_STATES or _CheckForOfflineNodes(nodes, inst): |
312 |
logging.info("Skipping instance '%s' because it is in a helpless state"
|
313 |
" or has offline secondaries", name)
|
314 |
continue
|
315 |
|
316 |
job.append(opcodes.OpInstanceActivateDisks(instance_name=name)) |
317 |
|
318 |
if job:
|
319 |
job_id = cli.SendJob(job, cl=cl) |
320 |
|
321 |
try:
|
322 |
cli.PollJob(job_id, cl=cl, feedback_fn=logging.debug) |
323 |
except Exception: # pylint: disable=W0703 |
324 |
logging.exception("Error while activating disks")
|
325 |
|
326 |
|
327 |
def IsRapiResponding(hostname): |
328 |
"""Connects to RAPI port and does a simple test.
|
329 |
|
330 |
Connects to RAPI port of hostname and does a simple test. At this time, the
|
331 |
test is GetVersion.
|
332 |
|
333 |
@type hostname: string
|
334 |
@param hostname: hostname of the node to connect to.
|
335 |
@rtype: bool
|
336 |
@return: Whether RAPI is working properly
|
337 |
|
338 |
"""
|
339 |
curl_config = rapi.client.GenericCurlConfig() |
340 |
rapi_client = rapi.client.GanetiRapiClient(hostname, |
341 |
curl_config_fn=curl_config) |
342 |
try:
|
343 |
master_version = rapi_client.GetVersion() |
344 |
except rapi.client.CertificateError, err:
|
345 |
logging.warning("RAPI certificate error: %s", err)
|
346 |
return False |
347 |
except rapi.client.GanetiApiError, err:
|
348 |
logging.warning("RAPI error: %s", err)
|
349 |
return False |
350 |
else:
|
351 |
logging.debug("Reported RAPI version %s", master_version)
|
352 |
return master_version == constants.RAPI_VERSION
|
353 |
|
354 |
|
355 |
def ParseOptions(): |
356 |
"""Parse the command line options.
|
357 |
|
358 |
@return: (options, args) as from OptionParser.parse_args()
|
359 |
|
360 |
"""
|
361 |
parser = OptionParser(description="Ganeti cluster watcher",
|
362 |
usage="%prog [-d]",
|
363 |
version="%%prog (ganeti) %s" %
|
364 |
constants.RELEASE_VERSION) |
365 |
|
366 |
parser.add_option(cli.DEBUG_OPT) |
367 |
parser.add_option(cli.NODEGROUP_OPT) |
368 |
parser.add_option("-A", "--job-age", dest="job_age", default=6 * 3600, |
369 |
help="Autoarchive jobs older than this age (default"
|
370 |
" 6 hours)")
|
371 |
parser.add_option("--ignore-pause", dest="ignore_pause", default=False, |
372 |
action="store_true", help="Ignore cluster pause setting") |
373 |
parser.add_option("--wait-children", dest="wait_children", |
374 |
action="store_true", help="Wait for child processes") |
375 |
parser.add_option("--no-wait-children", dest="wait_children", |
376 |
action="store_false",
|
377 |
help="Don't wait for child processes")
|
378 |
# See optparse documentation for why default values are not set by options
|
379 |
parser.set_defaults(wait_children=True)
|
380 |
options, args = parser.parse_args() |
381 |
options.job_age = cli.ParseTimespec(options.job_age) |
382 |
|
383 |
if args:
|
384 |
parser.error("No arguments expected")
|
385 |
|
386 |
return (options, args)
|
387 |
|
388 |
|
389 |
def _WriteInstanceStatus(filename, data): |
390 |
"""Writes the per-group instance status file.
|
391 |
|
392 |
The entries are sorted.
|
393 |
|
394 |
@type filename: string
|
395 |
@param filename: Path to instance status file
|
396 |
@type data: list of tuple; (instance name as string, status as string)
|
397 |
@param data: Instance name and status
|
398 |
|
399 |
"""
|
400 |
logging.debug("Updating instance status file '%s' with %s instances",
|
401 |
filename, len(data))
|
402 |
|
403 |
utils.WriteFile(filename, |
404 |
data="".join(map(compat.partial(operator.mod, "%s %s\n"), |
405 |
sorted(data))))
|
406 |
|
407 |
|
408 |
def _UpdateInstanceStatus(filename, instances): |
409 |
"""Writes an instance status file from L{Instance} objects.
|
410 |
|
411 |
@type filename: string
|
412 |
@param filename: Path to status file
|
413 |
@type instances: list of L{Instance}
|
414 |
|
415 |
"""
|
416 |
_WriteInstanceStatus(filename, [(inst.name, inst.status) |
417 |
for inst in instances]) |
418 |
|
419 |
|
420 |
def _ReadInstanceStatus(filename): |
421 |
"""Reads an instance status file.
|
422 |
|
423 |
@type filename: string
|
424 |
@param filename: Path to status file
|
425 |
@rtype: tuple; (None or number, list of lists containing instance name and
|
426 |
status)
|
427 |
@return: File's mtime and instance status contained in the file; mtime is
|
428 |
C{None} if file can't be read
|
429 |
|
430 |
"""
|
431 |
logging.debug("Reading per-group instance status from '%s'", filename)
|
432 |
|
433 |
statcb = utils.FileStatHelper() |
434 |
try:
|
435 |
content = utils.ReadFile(filename, preread=statcb) |
436 |
except EnvironmentError, err: |
437 |
if err.errno == errno.ENOENT:
|
438 |
logging.error("Can't read '%s', does not exist (yet)", filename)
|
439 |
else:
|
440 |
logging.exception("Unable to read '%s', ignoring", filename)
|
441 |
return (None, None) |
442 |
else:
|
443 |
return (statcb.st.st_mtime, [line.split(None, 1) |
444 |
for line in content.splitlines()]) |
445 |
|
446 |
|
447 |
def _MergeInstanceStatus(filename, pergroup_filename, groups): |
448 |
"""Merges all per-group instance status files into a global one.
|
449 |
|
450 |
@type filename: string
|
451 |
@param filename: Path to global instance status file
|
452 |
@type pergroup_filename: string
|
453 |
@param pergroup_filename: Path to per-group status files, must contain "%s"
|
454 |
to be replaced with group UUID
|
455 |
@type groups: sequence
|
456 |
@param groups: UUIDs of known groups
|
457 |
|
458 |
"""
|
459 |
# Lock global status file in exclusive mode
|
460 |
lock = utils.FileLock.Open(filename) |
461 |
try:
|
462 |
lock.Exclusive(blocking=True, timeout=INSTANCE_STATUS_LOCK_TIMEOUT)
|
463 |
except errors.LockError, err:
|
464 |
# All per-group processes will lock and update the file. None of them
|
465 |
# should take longer than 10 seconds (the value of
|
466 |
# INSTANCE_STATUS_LOCK_TIMEOUT).
|
467 |
logging.error("Can't acquire lock on instance status file '%s', not"
|
468 |
" updating: %s", filename, err)
|
469 |
return
|
470 |
|
471 |
logging.debug("Acquired exclusive lock on '%s'", filename)
|
472 |
|
473 |
data = {} |
474 |
|
475 |
# Load instance status from all groups
|
476 |
for group_uuid in groups: |
477 |
(mtime, instdata) = _ReadInstanceStatus(pergroup_filename % group_uuid) |
478 |
|
479 |
if mtime is not None: |
480 |
for (instance_name, status) in instdata: |
481 |
data.setdefault(instance_name, []).append((mtime, status)) |
482 |
|
483 |
# Select last update based on file mtime
|
484 |
inststatus = [(instance_name, sorted(status, reverse=True)[0][1]) |
485 |
for (instance_name, status) in data.items()] |
486 |
|
487 |
# Write the global status file. Don't touch file after it's been
|
488 |
# updated--there is no lock anymore.
|
489 |
_WriteInstanceStatus(filename, inststatus) |
490 |
|
491 |
|
492 |
def GetLuxiClient(try_restart): |
493 |
"""Tries to connect to the master daemon.
|
494 |
|
495 |
@type try_restart: bool
|
496 |
@param try_restart: Whether to attempt to restart the master daemon
|
497 |
|
498 |
"""
|
499 |
try:
|
500 |
return cli.GetClient()
|
501 |
except errors.OpPrereqError, err:
|
502 |
# this is, from cli.GetClient, a not-master case
|
503 |
raise NotMasterError("Not on master node (%s)" % err) |
504 |
|
505 |
except luxi.NoMasterError, err:
|
506 |
if not try_restart: |
507 |
raise
|
508 |
|
509 |
logging.warning("Master daemon seems to be down (%s), trying to restart",
|
510 |
err) |
511 |
|
512 |
if not utils.EnsureDaemon(constants.MASTERD): |
513 |
raise errors.GenericError("Can't start the master daemon") |
514 |
|
515 |
# Retry the connection
|
516 |
return cli.GetClient()
|
517 |
|
518 |
|
519 |
def _StartGroupChildren(cl, wait): |
520 |
"""Starts a new instance of the watcher for every node group.
|
521 |
|
522 |
"""
|
523 |
assert not compat.any(arg.startswith(cli.NODEGROUP_OPT_NAME) |
524 |
for arg in sys.argv) |
525 |
|
526 |
result = cl.QueryGroups([], ["name", "uuid"], False) |
527 |
|
528 |
children = [] |
529 |
|
530 |
for (idx, (name, uuid)) in enumerate(result): |
531 |
args = sys.argv + [cli.NODEGROUP_OPT_NAME, uuid] |
532 |
|
533 |
if idx > 0: |
534 |
# Let's not kill the system
|
535 |
time.sleep(CHILD_PROCESS_DELAY) |
536 |
|
537 |
logging.debug("Spawning child for group '%s' (%s), arguments %s",
|
538 |
name, uuid, args) |
539 |
|
540 |
try:
|
541 |
# TODO: Should utils.StartDaemon be used instead?
|
542 |
pid = os.spawnv(os.P_NOWAIT, args[0], args)
|
543 |
except Exception: # pylint: disable=W0703 |
544 |
logging.exception("Failed to start child for group '%s' (%s)",
|
545 |
name, uuid) |
546 |
else:
|
547 |
logging.debug("Started with PID %s", pid)
|
548 |
children.append(pid) |
549 |
|
550 |
if wait:
|
551 |
for pid in children: |
552 |
logging.debug("Waiting for child PID %s", pid)
|
553 |
try:
|
554 |
result = utils.RetryOnSignal(os.waitpid, pid, 0)
|
555 |
except EnvironmentError, err: |
556 |
result = str(err)
|
557 |
|
558 |
logging.debug("Child PID %s exited with status %s", pid, result)
|
559 |
|
560 |
|
561 |
def _ArchiveJobs(cl, age): |
562 |
"""Archives old jobs.
|
563 |
|
564 |
"""
|
565 |
(arch_count, left_count) = cl.AutoArchiveJobs(age) |
566 |
logging.debug("Archived %s jobs, left %s", arch_count, left_count)
|
567 |
|
568 |
|
569 |
def _CheckMaster(cl): |
570 |
"""Ensures current host is master node.
|
571 |
|
572 |
"""
|
573 |
(master, ) = cl.QueryConfigValues(["master_node"])
|
574 |
if master != netutils.Hostname.GetSysName():
|
575 |
raise NotMasterError("This is not the master node") |
576 |
|
577 |
|
578 |
@UsesRapiClient
|
579 |
def _GlobalWatcher(opts): |
580 |
"""Main function for global watcher.
|
581 |
|
582 |
At the end child processes are spawned for every node group.
|
583 |
|
584 |
"""
|
585 |
StartNodeDaemons() |
586 |
RunWatcherHooks() |
587 |
|
588 |
# Run node maintenance in all cases, even if master, so that old masters can
|
589 |
# be properly cleaned up
|
590 |
if nodemaint.NodeMaintenance.ShouldRun(): # pylint: disable=E0602 |
591 |
nodemaint.NodeMaintenance().Exec() # pylint: disable=E0602
|
592 |
|
593 |
try:
|
594 |
client = GetLuxiClient(True)
|
595 |
except NotMasterError:
|
596 |
# Don't proceed on non-master nodes
|
597 |
return constants.EXIT_SUCCESS
|
598 |
|
599 |
# we are on master now
|
600 |
utils.EnsureDaemon(constants.RAPI) |
601 |
|
602 |
# If RAPI isn't responding to queries, try one restart
|
603 |
logging.debug("Attempting to talk to remote API on %s",
|
604 |
constants.IP4_ADDRESS_LOCALHOST) |
605 |
if not IsRapiResponding(constants.IP4_ADDRESS_LOCALHOST): |
606 |
logging.warning("Couldn't get answer from remote API, restaring daemon")
|
607 |
utils.StopDaemon(constants.RAPI) |
608 |
utils.EnsureDaemon(constants.RAPI) |
609 |
logging.debug("Second attempt to talk to remote API")
|
610 |
if not IsRapiResponding(constants.IP4_ADDRESS_LOCALHOST): |
611 |
logging.fatal("RAPI is not responding")
|
612 |
logging.debug("Successfully talked to remote API")
|
613 |
|
614 |
_CheckMaster(client) |
615 |
_ArchiveJobs(client, opts.job_age) |
616 |
|
617 |
# Spawn child processes for all node groups
|
618 |
_StartGroupChildren(client, opts.wait_children) |
619 |
|
620 |
return constants.EXIT_SUCCESS
|
621 |
|
622 |
|
623 |
def _GetGroupData(cl, uuid): |
624 |
"""Retrieves instances and nodes per node group.
|
625 |
|
626 |
"""
|
627 |
job = [ |
628 |
# Get all primary instances in group
|
629 |
opcodes.OpQuery(what=constants.QR_INSTANCE, |
630 |
fields=["name", "status", "admin_state", "snodes", |
631 |
"pnode.group.uuid", "snodes.group.uuid"], |
632 |
qfilter=[qlang.OP_EQUAL, "pnode.group.uuid", uuid],
|
633 |
use_locking=True),
|
634 |
|
635 |
# Get all nodes in group
|
636 |
opcodes.OpQuery(what=constants.QR_NODE, |
637 |
fields=["name", "bootid", "offline"], |
638 |
qfilter=[qlang.OP_EQUAL, "group.uuid", uuid],
|
639 |
use_locking=True),
|
640 |
] |
641 |
|
642 |
job_id = cl.SubmitJob(job) |
643 |
results = map(objects.QueryResponse.FromDict,
|
644 |
cli.PollJob(job_id, cl=cl, feedback_fn=logging.debug)) |
645 |
cl.ArchiveJob(job_id) |
646 |
|
647 |
results_data = map(operator.attrgetter("data"), results) |
648 |
|
649 |
# Ensure results are tuples with two values
|
650 |
assert compat.all(map(ht.TListOf(ht.TListOf(ht.TIsLength(2))), results_data)) |
651 |
|
652 |
# Extract values ignoring result status
|
653 |
(raw_instances, raw_nodes) = [[map(compat.snd, values)
|
654 |
for values in res] |
655 |
for res in results_data] |
656 |
|
657 |
secondaries = {} |
658 |
instances = [] |
659 |
|
660 |
# Load all instances
|
661 |
for (name, status, autostart, snodes, pnode_group_uuid,
|
662 |
snodes_group_uuid) in raw_instances:
|
663 |
if snodes and set([pnode_group_uuid]) != set(snodes_group_uuid): |
664 |
logging.error("Ignoring split instance '%s', primary group %s, secondary"
|
665 |
" groups %s", name, pnode_group_uuid,
|
666 |
utils.CommaJoin(snodes_group_uuid)) |
667 |
else:
|
668 |
instances.append(Instance(name, status, autostart, snodes)) |
669 |
|
670 |
for node in snodes: |
671 |
secondaries.setdefault(node, set()).add(name)
|
672 |
|
673 |
# Load all nodes
|
674 |
nodes = [Node(name, bootid, offline, secondaries.get(name, set()))
|
675 |
for (name, bootid, offline) in raw_nodes] |
676 |
|
677 |
return (dict((node.name, node) for node in nodes), |
678 |
dict((inst.name, inst) for inst in instances)) |
679 |
|
680 |
|
681 |
def _LoadKnownGroups(): |
682 |
"""Returns a list of all node groups known by L{ssconf}.
|
683 |
|
684 |
"""
|
685 |
groups = ssconf.SimpleStore().GetNodegroupList() |
686 |
|
687 |
result = list(line.split(None, 1)[0] for line in groups |
688 |
if line.strip())
|
689 |
|
690 |
if not compat.all(map(utils.UUID_RE.match, result)): |
691 |
raise errors.GenericError("Ssconf contains invalid group UUID") |
692 |
|
693 |
return result
|
694 |
|
695 |
|
696 |
def _GroupWatcher(opts): |
697 |
"""Main function for per-group watcher process.
|
698 |
|
699 |
"""
|
700 |
group_uuid = opts.nodegroup.lower() |
701 |
|
702 |
if not utils.UUID_RE.match(group_uuid): |
703 |
raise errors.GenericError("Node group parameter (%s) must be given a UUID," |
704 |
" got '%s'" %
|
705 |
(cli.NODEGROUP_OPT_NAME, group_uuid)) |
706 |
|
707 |
logging.info("Watcher for node group '%s'", group_uuid)
|
708 |
|
709 |
known_groups = _LoadKnownGroups() |
710 |
|
711 |
# Check if node group is known
|
712 |
if group_uuid not in known_groups: |
713 |
raise errors.GenericError("Node group '%s' is not known by ssconf" % |
714 |
group_uuid) |
715 |
|
716 |
# Group UUID has been verified and should not contain any dangerous
|
717 |
# characters
|
718 |
state_path = pathutils.WATCHER_GROUP_STATE_FILE % group_uuid |
719 |
inst_status_path = pathutils.WATCHER_GROUP_INSTANCE_STATUS_FILE % group_uuid |
720 |
|
721 |
logging.debug("Using state file %s", state_path)
|
722 |
|
723 |
# Global watcher
|
724 |
statefile = state.OpenStateFile(state_path) # pylint: disable=E0602
|
725 |
if not statefile: |
726 |
return constants.EXIT_FAILURE
|
727 |
|
728 |
notepad = state.WatcherState(statefile) # pylint: disable=E0602
|
729 |
try:
|
730 |
# Connect to master daemon
|
731 |
client = GetLuxiClient(False)
|
732 |
|
733 |
_CheckMaster(client) |
734 |
|
735 |
(nodes, instances) = _GetGroupData(client, group_uuid) |
736 |
|
737 |
# Update per-group instance status file
|
738 |
_UpdateInstanceStatus(inst_status_path, instances.values()) |
739 |
|
740 |
_MergeInstanceStatus(pathutils.INSTANCE_STATUS_FILE, |
741 |
pathutils.WATCHER_GROUP_INSTANCE_STATUS_FILE, |
742 |
known_groups) |
743 |
|
744 |
started = _CheckInstances(client, notepad, instances) |
745 |
_CheckDisks(client, notepad, nodes, instances, started) |
746 |
_VerifyDisks(client, group_uuid, nodes, instances) |
747 |
except Exception, err: |
748 |
logging.info("Not updating status file due to failure: %s", err)
|
749 |
raise
|
750 |
else:
|
751 |
# Save changes for next run
|
752 |
notepad.Save(state_path) |
753 |
|
754 |
return constants.EXIT_SUCCESS
|
755 |
|
756 |
|
757 |
def Main(): |
758 |
"""Main function.
|
759 |
|
760 |
"""
|
761 |
(options, _) = ParseOptions() |
762 |
|
763 |
utils.SetupLogging(pathutils.LOG_WATCHER, sys.argv[0],
|
764 |
debug=options.debug, stderr_logging=options.debug) |
765 |
|
766 |
if ShouldPause() and not options.ignore_pause: |
767 |
logging.debug("Pause has been set, exiting")
|
768 |
return constants.EXIT_SUCCESS
|
769 |
|
770 |
# Try to acquire global watcher lock in shared mode
|
771 |
lock = utils.FileLock.Open(pathutils.WATCHER_LOCK_FILE) |
772 |
try:
|
773 |
lock.Shared(blocking=False)
|
774 |
except (EnvironmentError, errors.LockError), err: |
775 |
logging.error("Can't acquire lock on %s: %s",
|
776 |
pathutils.WATCHER_LOCK_FILE, err) |
777 |
return constants.EXIT_SUCCESS
|
778 |
|
779 |
if options.nodegroup is None: |
780 |
fn = _GlobalWatcher |
781 |
else:
|
782 |
# Per-nodegroup watcher
|
783 |
fn = _GroupWatcher |
784 |
|
785 |
try:
|
786 |
return fn(options)
|
787 |
except (SystemExit, KeyboardInterrupt): |
788 |
raise
|
789 |
except NotMasterError:
|
790 |
logging.debug("Not master, exiting")
|
791 |
return constants.EXIT_NOTMASTER
|
792 |
except errors.ResolverError, err:
|
793 |
logging.error("Cannot resolve hostname '%s', exiting", err.args[0]) |
794 |
return constants.EXIT_NODESETUP_ERROR
|
795 |
except errors.JobQueueFull:
|
796 |
logging.error("Job queue is full, can't query cluster state")
|
797 |
except errors.JobQueueDrainError:
|
798 |
logging.error("Job queue is drained, can't maintain cluster state")
|
799 |
except Exception, err: |
800 |
logging.exception(str(err))
|
801 |
return constants.EXIT_FAILURE
|
802 |
|
803 |
return constants.EXIT_SUCCESS
|