3 # Reboot specified nodes
9 import errno, time, traceback
12 import threading, popen2
18 from subprocess import PIPE, Popen
19 import ssh.pxssh as pxssh
20 import ssh.pexpect as pexpect
23 # Use our versions of telnetlib and pyssh
24 sys.path.insert(0, os.path.dirname(sys.argv[0]))
26 sys.path.insert(0, os.path.dirname(sys.argv[0]) + "/pyssh")
32 # Event class ID from pcu events
33 #NODE_POWER_CONTROL = 3
36 #MONITOR_USER_ID = 11142
39 logger = logging.getLogger("monitor")
43 class ExceptionNoTransport(Exception): pass
44 class ExceptionNotFound(Exception): pass
45 class ExceptionPassword(Exception): pass
46 class ExceptionTimeout(Exception): pass
47 class ExceptionPrompt(Exception): pass
48 class ExceptionSequence(Exception): pass
49 class ExceptionReset(Exception): pass
50 class ExceptionPort(Exception): pass
51 class ExceptionUsername(Exception): pass
53 def telnet_answer(telnet, expected, buffer):
56 output = telnet.read_until(expected, TELNET_TIMEOUT)
58 # logger.debug(output)
59 if output.find(expected) == -1:
60 raise ExceptionNotFound, "'%s' not found" % expected
62 telnet.write(buffer + "\r\n")
65 # PCU has model, host, preferred-port, user, passwd,
67 # This is an object derived directly form the PLCAPI DB fields
69 def __init__(self, plc_pcu_dict):
70 for field in ['username', 'password', 'site_id',
73 'node_ids', 'ports', ]:
74 if field in plc_pcu_dict:
75 self.__setattr__(field, plc_pcu_dict[field])
77 raise Exception("No such field %s in PCU object" % field)
79 # These are the convenience functions build around the PCU object.
81 def __init__(self, plc_pcu_dict):
82 PCU.__init__(self, plc_pcu_dict)
83 self.host = self.pcu_name()
86 if self.hostname is not None and self.hostname is not "":
88 elif self.ip is not None and self.ip is not "":
93 def nodeidToPort(self, node_id):
94 if node_id in self.node_ids:
95 for i in range(0, len(self.node_ids)):
96 if node_id == self.node_ids[i]:
99 raise Exception("No such Node ID: %d" % node_id)
101 # This class captures the observed pcu records from FindBadPCUs.py
103 def __init__(self, pcu_record_dict):
104 for field in ['nodenames', 'portstatus',
107 if field in pcu_record_dict:
108 if field == "reboot":
109 self.__setattr__("reboot_str", pcu_record_dict[field])
111 self.__setattr__(field, pcu_record_dict[field])
113 raise Exception("No such field %s in pcu record dict" % field)
123 def __init__(self, type, verbose):
125 self.verbose = verbose
126 self.transport = None
128 def open(self, host, username=None, password=None, prompt="User Name"):
131 if self.type == self.TELNET:
132 transport = telnetlib.Telnet(host, timeout=self.TELNET_TIMEOUT)
133 transport.set_debuglevel(self.verbose)
134 if username is not None:
135 self.transport = transport
136 self.ifThenSend(prompt, username, ExceptionUsername)
138 elif self.type == self.SSH:
139 if username is not None:
140 transport = pyssh.Ssh(username, host)
141 transport.set_debuglevel(self.verbose)
143 # TODO: have an ssh set_debuglevel() also...
145 raise Exception("Username cannot be None for ssh transport.")
146 elif self.type == self.HTTP:
147 self.url = "http://%s:%d/" % (host,80)
148 uri = "%s:%d" % (host,80)
151 authinfo = urllib2.HTTPPasswordMgrWithDefaultRealm()
152 authinfo.add_password (None, uri, username, password)
153 authhandler = urllib2.HTTPBasicAuthHandler( authinfo )
155 transport = urllib2.build_opener(authhandler)
158 raise Exception("Unknown transport type: %s" % self.type)
160 self.transport = transport
164 if self.type == self.TELNET:
165 self.transport.close()
166 elif self.type == self.SSH:
167 self.transport.close()
168 elif self.type == self.HTTP:
171 raise Exception("Unknown transport type %s" % self.type)
172 self.transport = None
174 def sendHTTP(self, resource, data):
176 print "POSTing '%s' to %s" % (data,self.url + resource)
179 f = self.transport.open(self.url + resource ,data)
184 except urllib2.URLError,err:
185 logger.info('Could not open http connection', err)
186 return "http transport error"
190 def sendPassword(self, password, prompt=None):
191 if self.type == self.TELNET:
193 self.ifThenSend("Password", password, ExceptionPassword)
195 self.ifThenSend(prompt, password, ExceptionPassword)
196 elif self.type == self.SSH:
197 self.ifThenSend("password:", password, ExceptionPassword)
198 elif self.type == self.HTTP:
201 raise Exception("Unknown transport type: %s" % self.type)
203 def ifThenSend(self, expected, buffer, ErrorClass=ExceptionPrompt):
205 if self.transport != None:
206 output = self.transport.read_until(expected, self.TELNET_TIMEOUT)
207 if output.find(expected) == -1:
208 raise ErrorClass, "'%s' not found" % expected
210 self.transport.write(buffer + "\r\n")
212 raise ExceptionNoTransport("transport object is type None")
214 def ifElse(self, expected, ErrorClass):
216 self.transport.read_until(expected, self.TELNET_TIMEOUT)
218 raise ErrorClass("Could not find '%s' within timeout" % expected)
221 class PCUControl(Transport,PCUModel,PCURecord):
222 def __init__(self, plc_pcu_record, verbose, supported_ports=[]):
223 PCUModel.__init__(self, plc_pcu_record)
224 PCURecord.__init__(self, plc_pcu_record)
227 if '22' in supported_ports and self.portstatus['22'] == "open":
229 elif '23' in supported_ports and self.portstatus['23'] == "open":
230 type = Transport.TELNET
231 elif '80' in supported_ports and self.portstatus['80'] == "open":
232 type = Transport.HTTP
233 elif '443' in supported_ports and self.portstatus['443'] == "open":
234 type = Transport.HTTP
235 elif '5869' in supported_ports and self.portstatus['5869'] == "open":
236 # For DRAC cards. Racadm opens this port.
237 type = Transport.HTTP
238 elif '9100' in supported_ports and self.portstatus['9100'] == "open":
239 type = Transport.IPAL
240 elif '16992' in supported_ports and self.portstatus['16992'] == "open":
241 type = Transport.HTTP
243 raise ExceptionPort("Unsupported Port: No transport from open ports")
245 raise Exception("No Portstatus: No transport because no open ports")
246 Transport.__init__(self, type, verbose)
248 def run(self, node_port, dryrun):
249 """ This function is to be defined by the specific PCU instance. """
252 def reboot(self, node_port, dryrun):
254 return self.run(node_port, dryrun)
255 except ExceptionNotFound, err:
256 return "error: " + str(err)
257 except ExceptionPassword, err:
258 return "password exception: " + str(err)
259 except ExceptionTimeout, err:
260 return "timeout exception: " + str(err)
261 except ExceptionUsername, err:
262 return "exception: no username prompt: " + str(err)
263 except ExceptionSequence, err:
264 return "sequence error: " + str(err)
265 except ExceptionPrompt, err:
266 return "prompt exception: " + str(err)
267 except ExceptionPort, err:
268 return "no ports exception: " + str(err)
269 except socket.error, err:
270 return "socket error: timeout: " + str(err)
271 except EOFError, err:
273 logger.debug("reboot: EOF")
275 self.transport.close()
277 traceback.print_exc()
278 return "EOF connection reset" + str(err)
280 class IPAL(PCUControl):
282 This now uses a proprietary format for communicating with the PCU. I
283 prefer it to Telnet, and Web access, since it's much lighter weight
284 and, more importantly, IT WORKS!! HHAHHHAHAHAHAHAHA!
287 def format_msg(self, data, cmd):
288 esc = chr(int('1b',16))
289 return "%c%s%c%s%c" % (esc, self.password, esc, data, cmd) # esc, 'q', chr(4))
291 def recv_noblock(self, s, count):
295 # TODO: make sleep backoff, before stopping.
297 ret = s.recv(count, socket.MSG_DONTWAIT)
298 except socket.error, e:
299 if e[0] == errno.EAGAIN:
300 raise Exception(e[1])
302 # TODO: not other exceptions.
306 def run(self, node_port, dryrun):
312 s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
315 s.connect((self.host, 9100))
316 except socket.error, e:
318 if e[0] == errno.ECONNREFUSED:
319 # cannot connect to remote host
320 raise Exception(e[1])
322 # TODO: what other conditions are there?
326 print "Checking status"
327 s.send(self.format_msg("", 'O'))
328 ret = self.recv_noblock(s, 8)
329 print "Current status is '%s'" % ret
332 raise Exception("Status returned 'another session already open' %s : %s" % (node_port, ret))
335 if node_port < len(ret):
336 status = ret[node_port]
344 raise Exception("Unknown status for PCU socket %s : %s" % (node_port, ret))
346 raise Exception("Mismatch between configured port and PCU status: %s %s" % (node_port, ret))
351 print "Pulsing %s" % node_port
352 s.send(self.format_msg("%s" % node_port, 'P'))
354 # NOTE: turn power on ; do not pulse the port.
355 print "Power was off, so turning on ..."
356 s.send(self.format_msg("%s" % node_port, 'E'))
357 #s.send(self.format_msg("%s" % node_port, 'P'))
359 print "Receiving response."
360 ret = self.recv_noblock(s, 8)
361 print "Current status is '%s'" % ret
363 if node_port < len(ret):
364 status = ret[node_port]
372 raise Exception("Unknown status for PCU socket %s : %s" % (node_port, ret))
374 raise Exception("Mismatch between configured port and PCU status: %s %s" % (node_port, ret))
379 return "Failed Power On"
384 # TELNET version of protocol...
385 # #self.open(self.host)
386 # ## XXX Some iPals require you to hit Enter a few times first
387 # #self.ifThenSend("Password >", "\r\n\r\n", ExceptionNotFound)
389 # self.ifThenSend("Password >", self.password, ExceptionPassword)
390 # self.transport.write("\r\n\r\n")
391 # if not dryrun: # P# - Pulse relay
392 # print "node_port %s" % node_port
393 # self.ifThenSend("Enter >",
394 # "P7", # % node_port,
396 # print "send newlines"
397 # self.transport.write("\r\n\r\n")
398 # print "after new lines"
399 # # Get the next prompt
400 # print "wait for enter"
401 # self.ifElse("Enter >", ExceptionTimeout)
406 class APCEurope(PCUControl):
407 def run(self, node_port, dryrun):
408 self.open(self.host, self.username)
409 self.sendPassword(self.password)
411 self.ifThenSend("\r\n> ", "1", ExceptionPassword)
412 self.ifThenSend("\r\n> ", "2")
413 self.ifThenSend("\r\n> ", str(node_port))
414 # 3- Immediate Reboot
415 self.ifThenSend("\r\n> ", "3")
418 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
422 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
423 "", ExceptionSequence)
424 self.ifThenSend("Press <ENTER> to continue...", "", ExceptionSequence)
429 class APCBrazil(PCUControl):
430 def run(self, node_port, dryrun):
431 self.open(self.host, self.username)
432 self.sendPassword(self.password)
434 self.ifThenSend("\r\n> ", "1", ExceptionPassword)
435 self.ifThenSend("\r\n> ", str(node_port))
436 # 4- Immediate Reboot
437 self.ifThenSend("\r\n> ", "4")
440 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
444 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
445 "", ExceptionSequence)
446 self.ifThenSend("Press <ENTER> to continue...", "", ExceptionSequence)
451 class APCBerlin(PCUControl):
452 def run(self, node_port, dryrun):
453 self.open(self.host, self.username)
454 self.sendPassword(self.password)
456 self.ifThenSend("\r\n> ", "1", ExceptionPassword)
457 self.ifThenSend("\r\n> ", "2")
458 self.ifThenSend("\r\n> ", "1")
459 self.ifThenSend("\r\n> ", str(node_port))
460 # 3- Immediate Reboot
461 self.ifThenSend("\r\n> ", "3")
464 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
468 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
469 "", ExceptionSequence)
470 self.ifThenSend("Press <ENTER> to continue...", "", ExceptionSequence)
475 class APCFolsom(PCUControl):
476 def run(self, node_port, dryrun):
477 self.open(self.host, self.username)
478 self.sendPassword(self.password)
480 self.ifThenSend("\r\n> ", "1", ExceptionPassword)
481 self.ifThenSend("\r\n> ", "2")
482 self.ifThenSend("\r\n> ", "1")
483 self.ifThenSend("\r\n> ", str(node_port))
484 self.ifThenSend("\r\n> ", "1")
486 # 3- Immediate Reboot
487 self.ifThenSend("\r\n> ", "3")
490 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
494 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
495 "", ExceptionSequence)
496 self.ifThenSend("Press <ENTER> to continue...", "", ExceptionSequence)
501 class APCMaster(PCUControl):
502 def run(self, node_port, dryrun):
503 print "Rebooting %s" % self.host
504 self.open(self.host, self.username)
505 self.sendPassword(self.password)
508 self.ifThenSend("\r\n> ", "1", ExceptionPassword)
509 # 3- Outlet Control/Config
510 self.ifThenSend("\r\n> ", "3")
512 self.ifThenSend("\r\n> ", str(node_port))
514 self.ifThenSend("\r\n> ", "1")
515 # 3- Immediate Reboot
516 self.ifThenSend("\r\n> ", "3")
519 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
523 self.ifThenSend("Enter 'YES' to continue or <ENTER> to cancel",
524 "", ExceptionSequence)
525 self.ifThenSend("Press <ENTER> to continue...", "", ExceptionSequence)
530 class APC(PCUControl):
531 def __init__(self, plc_pcu_record, verbose):
532 PCUControl.__init__(self, plc_pcu_record, verbose)
534 self.master = APCMaster(plc_pcu_record, verbose)
535 self.folsom = APCFolsom(plc_pcu_record, verbose)
536 self.europe = APCEurope(plc_pcu_record, verbose)
538 def run(self, node_port, dryrun):
542 for pcu in [self.master, self.europe, self.folsom]:
545 print "-*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*_*"
548 time.sleep(sleep_time)
549 ret = pcu.reboot(node_port, dryrun)
550 except ExceptionSequence, err:
556 return "Unknown reboot sequence for APC PCU"
560 class IntelAMT(PCUControl):
561 def run(self, node_port, dryrun):
565 cmd_str = "IntelAMTSDK/Samples/RemoteControl/remoteControl"
568 # NOTE: -p checks the power state of the host.
569 # TODO: parse the output to find out if it's ok or not.
570 cmd_str += " -p http://%s:16992/RemoteControlService -user admin -pass '%s' " % (self.host, self.password )
572 cmd_str += " -A http://%s:16992/RemoteControlService -user admin -pass '%s' " % (self.host, self.password )
575 return cmd.system(cmd_str, self.TELNET_TIMEOUT)
577 class DRACRacAdm(PCUControl):
578 def run(self, node_port, dryrun):
580 print "trying racadm_reboot..."
581 racadm_reboot(self.host, self.username, self.password, node_port, dryrun)
585 class DRAC(PCUControl):
586 def run(self, node_port, dryrun):
587 self.open(self.host, self.username)
588 self.sendPassword(self.password)
590 print "logging in..."
591 self.transport.write("\r\n")
594 self.ifThenSend("[%s]#" % self.username, "getsysinfo")
597 self.ifThenSend("[%s]#" % self.username, "serveraction powercycle")
599 self.ifThenSend("[%s]#" % self.username, "exit")
604 class HPiLO(PCUControl):
605 def run(self, node_port, dryrun):
606 self.open(self.host, self.username)
607 self.sendPassword(self.password)
610 self.ifThenSend("</>hpiLO->", "cd system1")
612 # Reboot Outlet N (Y/N)?
614 self.ifThenSend("</system1>hpiLO->", "POWER")
617 self.ifThenSend("</system1>hpiLO->", "reset")
619 self.ifThenSend("</system1>hpiLO->", "exit")
625 class HPiLOHttps(PCUControl):
626 def run(self, node_port, dryrun):
629 locfg = soltesz.CMD()
630 cmd = "cmdhttps/locfg.pl -s %s -f %s -u %s -p '%s' | grep 'MESSAGE' | grep -v 'No error'" % (
631 self.host, "iloxml/Get_Network.xml",
632 self.username, self.password)
633 sout, serr = locfg.run_noexcept(cmd)
635 if sout.strip() != "":
636 print "sout: %s" % sout.strip()
640 locfg = soltesz.CMD()
641 cmd = "cmdhttps/locfg.pl -s %s -f %s -u %s -p '%s' | grep 'MESSAGE' | grep -v 'No error'" % (
642 self.host, "iloxml/Reset_Server.xml",
643 self.username, self.password)
644 sout, serr = locfg.run_noexcept(cmd)
646 if sout.strip() != "":
647 print "sout: %s" % sout.strip()
651 class BayTechAU(PCUControl):
652 def run(self, node_port, dryrun):
653 self.open(self.host, self.username, None, "Enter user name:")
654 self.sendPassword(self.password, "Enter Password:")
656 #self.ifThenSend("RPC-16>", "Status")
657 self.ifThenSend("RPC3-NC>", "Reboot %d" % node_port)
659 # Reboot Outlet N (Y/N)?
661 self.ifThenSend("(Y/N)?", "N")
663 self.ifThenSend("(Y/N)?", "Y")
664 self.ifThenSend("RPC3-NC>", "")
669 class BayTechGeorgeTown(PCUControl):
670 def run(self, node_port, dryrun):
671 self.open(self.host, self.username, None, "Enter user name:")
672 self.sendPassword(self.password, "Enter Password:")
674 #self.ifThenSend("RPC-16>", "Status")
676 self.ifThenSend("RPC-16>", "Reboot %d" % node_port)
678 # Reboot Outlet N (Y/N)?
680 self.ifThenSend("(Y/N)?", "N")
682 self.ifThenSend("(Y/N)?", "Y")
683 self.ifThenSend("RPC-16>", "")
688 class BayTechCtrlCUnibe(PCUControl):
690 For some reason, these units let you log in fine, but they hang
691 indefinitely, unless you send a Ctrl-C after the password. No idea
694 def run(self, node_port, dryrun):
695 print "BayTechCtrlC %s" % self.host
697 ssh_options="-o StrictHostKeyChecking=no -o PasswordAuthentication=yes -o PubkeyAuthentication=no"
699 if not s.login(self.host, self.username, self.password, ssh_options):
700 raise ExceptionPassword("Invalid Password")
701 # Otherwise, the login succeeded.
703 # Send a ctrl-c to the remote process.
704 print "sending ctrl-c"
707 # Control Outlets (5 ,1).........5
709 index = s.expect(["Enter Request :"])
714 index = s.expect(["DS-RPC>", "Enter user name:"])
716 s.send(self.username + "\r\n")
717 index = s.expect(["DS-RPC>"])
720 print "Reboot %d" % node_port
721 s.send("Reboot %d\r\n" % node_port)
723 index = s.expect(["(Y/N)?"])
732 #index = s.expect(["DS-RPC>"])
733 #print "got prompt back"
738 raise ExceptionPrompt("EOF before 'Enter Request' Prompt")
739 except pexpect.TIMEOUT:
740 raise ExceptionPrompt("Timeout before 'Enter Request' Prompt")
744 class BayTechCtrlC(PCUControl):
746 For some reason, these units let you log in fine, but they hang
747 indefinitely, unless you send a Ctrl-C after the password. No idea
750 def run(self, node_port, dryrun):
751 print "BayTechCtrlC %s" % self.host
753 ssh_options="-o StrictHostKeyChecking=no -o PasswordAuthentication=yes -o PubkeyAuthentication=no"
755 if not s.login(self.host, self.username, self.password, ssh_options):
756 raise ExceptionPassword("Invalid Password")
757 # Otherwise, the login succeeded.
759 # Send a ctrl-c to the remote process.
760 print "sending ctrl-c"
763 # Control Outlets (5 ,1).........5
765 index = s.expect(["Enter Request :"])
770 index = s.expect(["DS-RPC>", "Enter user name:"])
772 print "sending username"
773 s.send(self.username + "\r\n")
774 index = s.expect(["DS-RPC>"])
777 print "Reboot %d" % node_port
778 s.send("Reboot %d\r\n" % node_port)
780 index = s.expect(["(Y/N)?"])
789 index = s.expect(["DS-RPC>"])
790 #print "got prompt back"
795 raise ExceptionPrompt("EOF before 'Enter Request' Prompt")
796 except pexpect.TIMEOUT:
797 raise ExceptionPrompt("Timeout before Prompt")
801 class BayTech(PCUControl):
802 def run(self, node_port, dryrun):
803 self.open(self.host, self.username)
804 self.sendPassword(self.password)
806 # Control Outlets (5 ,1).........5
807 self.ifThenSend("Enter Request :", "5")
811 self.ifThenSend("DS-RPC>", "Reboot %d" % node_port, ExceptionNotFound)
812 except ExceptionNotFound, msg:
813 # one machine is configured to ask for a username,
814 # even after login...
815 print "msg: %s" % msg
816 self.transport.write(self.username + "\r\n")
817 self.ifThenSend("DS-RPC>", "Reboot %d" % node_port)
819 # Reboot Outlet N (Y/N)?
821 self.ifThenSend("(Y/N)?", "N")
823 self.ifThenSend("(Y/N)?", "Y")
824 self.ifThenSend("DS-RPC>", "")
829 class WTIIPS4(PCUControl):
830 def run(self, node_port, dryrun):
832 self.sendPassword(self.password, "Enter Password:")
834 self.ifThenSend("IPS> ", "/Boot %s" % node_port)
836 self.ifThenSend("Sure? (Y/N): ", "N")
838 self.ifThenSend("Sure? (Y/N): ", "Y")
840 self.ifThenSend("IPS> ", "")
845 class ePowerSwitchGood(PCUControl):
847 # The old code used Python's HTTPPasswordMgrWithDefaultRealm()
848 # For some reason this both doesn't work and in some cases, actually
849 # hangs the PCU. Definitely not what we want.
851 # The code below is much simpler. Just letting things fail first,
852 # and then, trying again with authentication string in the header.
854 def run(self, node_port, dryrun):
855 self.transport = None
856 self.url = "http://%s:%d/" % (self.host,80)
857 uri = "%s:%d" % (self.host,80)
859 req = urllib2.Request(self.url)
861 handle = urllib2.urlopen(req)
863 # NOTE: this is expected to fail initially
870 return "ERROR: not protected by HTTP authentication"
872 if not hasattr(e, 'code') or e.code != 401:
873 return "ERROR: failed for: %s" % str(e)
875 base64data = base64.encodestring("%s:%s" % (self.username, self.password))[:-1]
876 # NOTE: assuming basic realm authentication.
877 authheader = "Basic %s" % base64data
878 req.add_header("Authorization", authheader)
881 f = urllib2.urlopen(req)
883 # failing here means the User/passwd is wrong (hopefully)
884 raise ExceptionPassword("Incorrect username/password")
886 # NOTE: after verifying that the user/password is correct,
887 # actually reboot the given node.
890 data = urllib.urlencode({'P%d' % node_port : "r"})
891 req = urllib2.Request(self.url + "cmd.html")
892 req.add_header("Authorization", authheader)
893 # add data to handler,
894 f = urllib2.urlopen(req, data)
895 if self.verbose: print f.read()
897 import traceback; traceback.print_exc()
899 # fetch url one more time on cmd.html, econtrol.html or whatever.
902 if self.verbose: print f.read()
908 class ePowerSwitchOld(PCUControl):
909 def run(self, node_port, dryrun):
910 self.url = "http://%s:%d/" % (self.host,80)
911 uri = "%s:%d" % (self.host,80)
914 authinfo = urllib2.HTTPPasswordMgrWithDefaultRealm()
915 authinfo.add_password (None, uri, self.username, self.password)
916 authhandler = urllib2.HTTPBasicAuthHandler( authinfo )
918 # NOTE: it doesn't seem to matter whether this authinfo is here or not.
919 transport = urllib2.build_opener(authinfo)
920 f = transport.open(self.url)
921 if self.verbose: print f.read()
924 transport = urllib2.build_opener(authhandler)
925 f = transport.open(self.url + "cmd.html", "P%d=r" % node_port)
926 if self.verbose: print f.read()
931 class ePowerSwitch(PCUControl):
932 def run(self, node_port, dryrun):
933 self.url = "http://%s:%d/" % (self.host,80)
934 uri = "%s:%d" % (self.host,80)
936 # TODO: I'm still not sure what the deal is here.
937 # two independent calls appear to need to be made before the
938 # reboot will succeed. It doesn't seem to be possible to do
939 # this with a single call. I have no idea why.
942 authinfo = urllib2.HTTPPasswordMgrWithDefaultRealm()
943 authinfo.add_password (None, uri, self.username, self.password)
944 authhandler = urllib2.HTTPBasicAuthHandler( authinfo )
946 # NOTE: it doesn't seem to matter whether this authinfo is here or not.
947 transport = urllib2.build_opener()
948 f = transport.open(self.url + "elogin.html", "pwd=%s" % self.password)
949 if self.verbose: print f.read()
952 transport = urllib2.build_opener(authhandler)
953 f = transport.open(self.url + "econtrol.html", "P%d=r" % node_port)
954 if self.verbose: print f.read()
956 # data= "P%d=r" % node_port
957 #self.open(self.host, self.username, self.password)
958 #self.sendHTTP("elogin.html", "pwd=%s" % self.password)
959 #self.sendHTTP("econtrol.html", data)
960 #self.sendHTTP("cmd.html", data)
966 ### rebooting european BlackBox PSE boxes
967 # Thierry Parmentelat - May 11 2005
968 # tested on 4-ports models known as PSE505-FR
969 # uses http to POST a data 'P<port>=r'
970 # relies on basic authentication within http1.0
971 # first curl-based script was
972 # curl --http1.0 --basic --user <username>:<password> --data P<port>=r \
973 # http://<hostname>:<http_port>/cmd.html && echo OK
975 def bbpse_reboot (pcu_ip,username,password,port_in_pcu,http_port, dryrun):
979 url = "http://%s:%d/cmd.html" % (pcu_ip,http_port)
980 data= "P%d=r" % port_in_pcu
982 logger.debug("POSTing '%s' on %s" % (data,url))
984 authinfo = urllib2.HTTPPasswordMgrWithDefaultRealm()
985 uri = "%s:%d" % (pcu_ip,http_port)
986 authinfo.add_password (None, uri, username, password)
987 authhandler = urllib2.HTTPBasicAuthHandler( authinfo )
989 opener = urllib2.build_opener(authhandler)
990 urllib2.install_opener(opener)
996 f = urllib2.urlopen(url,data)
1003 except urllib2.URLError,err:
1004 logger.info('Could not open http connection', err)
1005 return "bbpse error"
1007 ### rebooting x10toggle based systems addressed by port
1008 # Marc E. Fiuczynski - May 31 2005
1009 # tested on 4-ports models known as PSE505-FR
1010 # uses ssh and password to login to an account
1011 # that will cause the system to be powercycled.
1013 def x10toggle_reboot(ip, username, password, port, dryrun):
1018 ssh = pyssh.Ssh(username, ip)
1022 telnet_answer(ssh, "password:", password)
1026 telnet_answer(ssh, "x10toggle>", "A%d" % port)
1029 output = ssh.close()
1031 logger.debug(output)
1034 except Exception, err:
1038 output = ssh.close()
1040 logger.debug(output)
1041 return errno.ETIMEDOUT
1043 ### rebooting Dell systems via RAC card
1044 # Marc E. Fiuczynski - June 01 2005
1045 # tested with David Lowenthal's itchy/scratchy nodes at UGA
1048 def runcmd(command, args, username, password, timeout = None):
1051 result_ready = threading.Condition()
1055 result_ready.acquire()
1059 result_ready.notify()
1060 result_ready.release()
1062 def do_command(command, username, password):
1065 # Popen4 is a popen-type class that combines stdout and stderr
1066 p = popen2.Popen4(command)
1068 # read all output data
1069 p.tochild.write("%s\n" % username)
1070 p.tochild.write("%s\n" % password)
1072 data = p.fromchild.read()
1075 # might get interrupted by a signal in poll() or waitpid()
1078 set_result((retval, data))
1081 if ex.errno == errno.EINTR:
1084 except Exception, ex:
1088 command = " ".join([command] + args)
1090 worker = threading.Thread(target = do_command, args = (command, username, password, ))
1091 worker.setDaemon(True)
1092 result_ready.acquire()
1094 result_ready.wait(timeout)
1096 if result == [None]:
1097 raise Exception, "command timed-out: '%s'" % command
1099 result_ready.release()
1102 if isinstance(result, Exception):
1105 (retval, data) = result
1106 if os.WIFEXITED(retval) and os.WEXITSTATUS(retval) == 0:
1109 out = "system command ('%s') " % command
1110 if os.WIFEXITED(retval):
1111 out += "failed, rc = %d" % os.WEXITSTATUS(retval)
1113 out += "killed by signal %d" % os.WTERMSIG(retval)
1115 out += "; output follows:\n" + data
1116 raise Exception, out
1118 def racadm_reboot(host, username, password, port, dryrun):
1121 ip = socket.gethostbyname(host)
1123 cmd = "/usr/sbin/racadm"
1126 output = runcmd(cmd, ["-r %s -i serveraction powercycle" % ip],
1129 output = runcmd(cmd, ["-r %s -i getsysinfo" % ip],
1132 print "RUNCMD: %s" % output
1134 logger.debug(output)
1137 except Exception, err:
1138 logger.debug("runcmd raised exception %s" % err)
1144 if pcu['hostname'] is not None and pcu['hostname'] is not "":
1145 return pcu['hostname']
1146 elif pcu['ip'] is not None and pcu['ip'] is not "":
1152 fb =soltesz.dbLoad("findbadpcus")
1154 def get_pcu_values(pcu_id):
1155 # TODO: obviously, this shouldn't be loaded each time...
1158 values = fb['nodes']["id_%s" % pcu_id]['values']
1164 def reboot(nodename):
1165 return reboot_policy(nodename, True, False)
1167 def reboot_policy(nodename, continue_probe, dryrun):
1170 pcu = plc.getpcu(nodename)
1172 logger.debug("no pcu for %s" % hostname)
1173 print "no pcu for %s" % hostname
1174 return False # "%s has no pcu" % nodename
1176 values = get_pcu_values(pcu['pcu_id'])
1178 logger.debug("No values for pcu probe %s" % hostname)
1179 print "No values for pcu probe %s" % hostname
1180 return False #"no info for pcu_id %s" % pcu['pcu_id']
1183 logger.debug("Trying PCU %s %s" % (pcu['hostname'], pcu['model']))
1185 ret = reboot_test(nodename, values, continue_probe, verbose, dryrun)
1194 def reboot_test(nodename, values, continue_probe, verbose, dryrun):
1198 # DataProbe iPal (many sites)
1199 if continue_probe and values['model'].find("IP-41x_IP-81x") >= 0:
1200 ipal = IPAL(values, verbose, ['23', '80', '9100'])
1201 rb_ret = ipal.reboot(values[nodename], dryrun)
1203 # APC Masterswitch (Berkeley)
1204 elif continue_probe and ( values['model'].find("AP79xx") >= 0 or \
1205 values['model'].find("Masterswitch") >= 0 ):
1208 # TODO: make a more robust version of APC
1209 if values['pcu_id'] in [1163,1055,1111,1231,1113,1127,1128,1148]:
1210 apc = APCEurope(values, verbose, ['22', '23'])
1211 rb_ret = apc.reboot(values[nodename], dryrun)
1213 elif values['pcu_id'] in [1110,86]:
1214 apc = APCBrazil(values, verbose, ['22', '23'])
1215 rb_ret = apc.reboot(values[nodename], dryrun)
1217 elif values['pcu_id'] in [1221,1225]:
1218 apc = APCBerlin(values, verbose, ['22', '23'])
1219 rb_ret = apc.reboot(values[nodename], dryrun)
1221 elif values['pcu_id'] in [1173,1221,1220]:
1222 apc = APCFolsom(values, verbose, ['22', '23'])
1223 rb_ret = apc.reboot(values[nodename], dryrun)
1226 apc = APCMaster(values, verbose, ['22', '23'])
1227 rb_ret = apc.reboot(values[nodename], dryrun)
1230 elif continue_probe and values['model'].find("DS4-RPC") >= 0:
1231 if values['pcu_id'] in [1237,1052,1209,1002,1008,1041,1013,1022]:
1232 # These require a 'ctrl-c' to be sent...
1233 baytech = BayTechCtrlC(values, verbose, ['22', '23'])
1234 rb_ret = baytech.reboot(values[nodename], dryrun)
1236 elif values['pcu_id'] in [93]:
1237 baytech = BayTechAU(values, verbose, ['22', '23'])
1238 rb_ret = baytech.reboot(values[nodename], dryrun)
1240 elif values['pcu_id'] in [1057]:
1241 # These require a 'ctrl-c' to be sent...
1242 baytech = BayTechCtrlCUnibe(values, verbose, ['22', '23'])
1243 rb_ret = baytech.reboot(values[nodename], dryrun)
1245 elif values['pcu_id'] in [1012]:
1246 # This pcu sometimes doesn't present the 'Username' prompt,
1247 # unless you immediately try again...
1249 baytech = BayTechGeorgeTown(values, verbose, ['22', '23'])
1250 rb_ret = baytech.reboot(values[nodename], dryrun)
1252 baytech = BayTechGeorgeTown(values, verbose, ['22', '23'])
1253 rb_ret = baytech.reboot(values[nodename], dryrun)
1255 baytech = BayTech(values, verbose, ['22', '23'])
1256 rb_ret = baytech.reboot(values[nodename], dryrun)
1259 elif continue_probe and values['model'].find("ilo") >= 0:
1261 hpilo = HPiLO(values, verbose, ['22'])
1262 rb_ret = hpilo.reboot(0, dryrun)
1264 hpilo = HPiLOHttps(values, verbose, ['443'])
1265 rb_ret = hpilo.reboot(0, dryrun)
1267 hpilo = HPiLOHttps(values, verbose, ['443'])
1268 rb_ret = hpilo.reboot(0, dryrun)
1271 elif continue_probe and values['model'].find("DRAC") >= 0:
1272 # TODO: I don't think DRACRacAdm will throw an exception for the
1273 # default method to catch...
1275 drac = DRACRacAdm(values, verbose, ['443', '5869'])
1276 rb_ret = drac.reboot(0, dryrun)
1278 drac = DRAC(values, verbose, ['22'])
1279 rb_ret = drac.reboot(0, dryrun)
1281 elif continue_probe and values['model'].find("WTI IPS-4") >= 0:
1282 wti = WTIIPS4(values, verbose, ['23'])
1283 rb_ret = wti.reboot(values[nodename], dryrun)
1285 elif continue_probe and values['model'].find("AMT") >= 0:
1286 amt = IntelAMT(values, verbose, ['16992'])
1287 rb_ret = amt.reboot(values[nodename], dryrun)
1289 # BlackBox PSExxx-xx (e.g. PSE505-FR)
1290 elif continue_probe and values['model'].find("ePowerSwitch") >=0:
1291 # TODO: allow a different port than http 80.
1292 if values['pcu_id'] in [1089, 1071, 1046, 1035, 1118]:
1293 eps = ePowerSwitchGood(values, verbose, ['80'])
1294 elif values['pcu_id'] in [1003]:
1297 eps = ePowerSwitch(values, verbose, ['80'])
1299 eps = ePowerSwitchGood(values, verbose, ['80'])
1301 rb_ret = eps.reboot(values[nodename], dryrun)
1303 elif continue_probe:
1304 rb_ret = "Unsupported_PCU"
1306 elif continue_probe == False:
1307 if 'portstatus' in values:
1314 except ExceptionPort, err:
1319 #elif continue_probe and values['protocol'] == "racadm" and \
1320 # values['model'] == "RAC":
1321 # rb_ret = racadm_reboot(pcu_name(values),
1322 # values['username'],
1323 # values['password'],
1328 logger.setLevel(logging.DEBUG)
1329 ch = logging.StreamHandler()
1330 ch.setLevel(logging.DEBUG)
1331 formatter = logging.Formatter('LOGGER - %(message)s')
1332 ch.setFormatter(formatter)
1333 logger.addHandler(ch)
1336 if "test" in sys.argv:
1341 for node in sys.argv[1:]:
1342 if node == "test": continue
1344 print "Rebooting %s" % node
1345 if reboot_policy(node, True, dryrun):
1349 except Exception, err:
1350 import traceback; traceback.print_exc()
1353 if __name__ == '__main__':
1355 logger = logging.getLogger("monitor")