11 from datetime import datetime,timedelta
16 from monitor.pcu import reboot
17 from monitor import config
18 from monitor.database import FindbadPCURecordSync, FindbadPCURecord
19 from monitor import util
20 from monitor.wrapper import plc
21 from nodequery import pcu_select
23 plc_lock = threading.Lock()
28 def nmap_portstatus(status):
30 l_nmap = status.split()
33 continue_probe = False
35 results = port.split('/')
36 ps[results[0]] = results[1]
37 if results[1] == "open":
39 return (ps, continue_probe)
44 #print "GetPCU from PLC %s" % pcuname
45 l_pcu = plc.GetPCUs({'pcu_id' : pcuname})
51 #print "GetPCU from file %s" % pcuname
52 l_pcus = database.dbLoad("pculist")
54 if i['pcu_id'] == pcuname:
63 def get_nodes(node_ids):
67 l_node = plc.getNodes(node_ids, ['hostname', 'last_contact', 'node_id', 'ports'])
70 plc_nodes = database.dbLoad("l_plcnodes")
72 if n['node_id'] in node_ids:
84 def get_plc_pcu_values(pcuname):
86 Try to contact PLC to get the PCU info.
87 If that fails, try a backup copy from the last run.
88 If that fails, return None
92 l_pcu = get_pcu(pcuname)
95 site_id = l_pcu['site_id']
96 node_ids = l_pcu['node_ids']
97 l_node = get_nodes(node_ids)
99 if l_node is not None:
101 values[node['hostname']] = node['ports'][0]
103 values['nodenames'] = [node['hostname'] for node in l_node]
105 # NOTE: this is for a dry run later. It doesn't matter which node.
106 values['node_id'] = l_node[0]['node_id']
114 def get_plc_site_values(site_id):
115 ### GET PLC SITE ######################
121 d_site = plc.getSites({'site_id': site_id}, ['max_slices', 'slice_ids', 'node_ids', 'login_base'])
126 plc_sites = database.dbLoad("l_plcsites")
127 for site in plc_sites:
128 if site['site_id'] == site_id:
132 traceback.print_exc()
137 if d_site is not None:
138 max_slices = d_site['max_slices']
139 num_slices = len(d_site['slice_ids'])
140 num_nodes = len(d_site['node_ids'])
141 loginbase = d_site['login_base']
142 values['plcsite'] = {'num_nodes' : num_nodes,
143 'max_slices' : max_slices,
144 'num_slices' : num_slices,
145 'login_base' : loginbase,
146 'status' : 'SUCCESS'}
154 def collectPingAndSSH(pcuname, cohash):
156 continue_probe = True
158 values = {'reboot' : 'novalue'}
159 ### GET PCU ######################
163 v = get_plc_pcu_values(pcuname)
164 if v['hostname'] is not None: v['hostname'] = v['hostname'].strip()
165 if v['ip'] is not None: v['ip'] = v['ip'].strip()
168 values['plc_pcu_stats'] = v
170 continue_probe = False
173 traceback.print_exc()
174 continue_probe = False
176 if b_except or not continue_probe: return (None, None, None)
179 #### COMPLETE ENTRY #######################
181 values['complete_entry'] = []
182 #if values['protocol'] is None or values['protocol'] is "":
183 # values['complete_entry'] += ["protocol"]
184 if values['plc_pcu_stats']['model'] is None or values['plc_pcu_stats']['model'] is "":
185 values['complete_entry'] += ["model"]
186 # Cannot continue due to this condition
187 continue_probe = False
189 if values['plc_pcu_stats']['password'] is None or values['plc_pcu_stats']['password'] is "":
190 values['complete_entry'] += ["password"]
191 # Cannot continue due to this condition
192 continue_probe = False
194 if len(values['complete_entry']) > 0:
195 continue_probe = False
197 if values['plc_pcu_stats']['hostname'] is None or values['plc_pcu_stats']['hostname'] is "":
198 values['complete_entry'] += ["hostname"]
199 if values['plc_pcu_stats']['ip'] is None or values['plc_pcu_stats']['ip'] is "":
200 values['complete_entry'] += ["ip"]
202 # If there are no nodes associated with this PCU, then we cannot continue.
203 if len(values['plc_pcu_stats']['node_ids']) == 0:
204 continue_probe = False
205 values['complete_entry'] += ['NoNodeIds']
207 #### DNS and IP MATCH #######################
208 if values['plc_pcu_stats']['hostname'] is not None and values['plc_pcu_stats']['hostname'] is not "" and \
209 values['plc_pcu_stats']['ip'] is not None and values['plc_pcu_stats']['ip'] is not "":
210 #print "Calling socket.gethostbyname(%s)" % values['hostname']
212 ipaddr = socket.gethostbyname(values['plc_pcu_stats']['hostname'])
213 if ipaddr == values['plc_pcu_stats']['ip']:
214 values['dnsmatch'] = "DNS-OK"
216 values['dnsmatch'] = "DNS-MISMATCH"
217 continue_probe = False
219 except Exception, err:
220 values['dnsmatch'] = "DNS-NOENTRY"
221 values['plc_pcu_stats']['hostname'] = values['plc_pcu_stats']['ip']
224 if values['plc_pcu_stats']['ip'] is not None and values['plc_pcu_stats']['ip'] is not "":
225 values['dnsmatch'] = "NOHOSTNAME"
226 values['plc_pcu_stats']['hostname'] = values['plc_pcu_stats']['ip']
228 values['dnsmatch'] = "NO-DNS-OR-IP"
229 values['plc_pcu_stats']['hostname'] = "No_entry_in_DB"
230 continue_probe = False
232 #### RUN NMAP ###############################
234 nmap = util.command.CMD()
235 (oval,eval) = nmap.run_noexcept("nmap -oG - -P0 -p22,23,80,443,5869,9100,16992 %s | grep Host:" % reboot.pcu_name(values['plc_pcu_stats']))
236 # NOTE: an empty / error value for oval, will still work.
237 (values['portstatus'], continue_probe) = nmap_portstatus(oval)
239 values['portstatus'] = None
242 ###### DRY RUN ############################
243 if 'node_ids' in values['plc_pcu_stats'] and len(values['plc_pcu_stats']['node_ids']) > 0:
244 rb_ret = reboot.reboot_test(values['plc_pcu_stats']['nodenames'][0], values, continue_probe, 1, True)
246 rb_ret = "Not_Run" # No nodes to test"
248 values['reboot'] = rb_ret
251 print "____________________________________"
254 print "____________________________________"
255 errors['traceback'] = traceback.format_exc()
256 print errors['traceback']
258 values['date_checked'] = time.time()
259 return (pcuname, values, errors)
261 def recordPingAndSSH(request, result):
265 (nodename, values, errors) = result
267 if values is not None:
268 pcu_id = int(nodename)
269 fbsync = FindbadPCURecordSync.findby_or_create(plc_pcuid=0,
270 if_new_set={'round': global_round})
271 global_round = fbsync.round
272 fbnodesync = FindbadPCURecordSync.findby_or_create(plc_pcuid=pcu_id,
273 if_new_set={'round' : global_round})
275 fbrec = FindbadPCURecord(
276 date_checked=datetime.fromtimestamp(values['date_checked']),
278 plc_pcu_stats=values['plc_pcu_stats'],
279 dns_status=values['dnsmatch'],
280 port_status=values['portstatus'],
281 entry_complete=" ".join(values['complete_entry']),
282 reboot_trial_status="%s" % values['reboot'],
284 fbnodesync.round = global_round
286 print "%d %s %s" % (count, nodename, values)
288 if errors is not None:
289 pcu_id = "id_%s" % nodename
290 errorState[pcu_id] = errors
291 database.dbDump("findbadpcu_errors", errorState)
293 # this will be called when an exception occurs within a thread
294 def handle_exception(request, result):
295 print "Exception occured in request %s" % request.requestID
297 print "Result: %s" % i
300 def checkAndRecordState(l_pcus, cohash):
304 tp = threadpool.ThreadPool(10)
306 # CREATE all the work requests
307 for pcuname in l_pcus:
308 pcu_id = int(pcuname)
309 fbnodesync = FindbadPCURecordSync.findby_or_create(plc_pcuid=pcu_id, if_new_set={'round' : 0})
311 node_round = fbnodesync.round
312 if node_round < global_round:
313 # recreate node stats when refreshed
314 #print "%s" % nodename
315 req = threadpool.WorkRequest(collectPingAndSSH, [pcuname, cohash], {},
316 None, recordPingAndSSH, handle_exception)
319 # We just skip it, since it's "up to date"
321 print "%d %s %s" % (count, pcu_id, node_round)
323 # WAIT while all the work requests are processed.
329 # if more than two hours
330 if time.time() - begin > (60*60*1):
331 print "findbadpcus.py has run out of time!!!!!!"
333 except KeyboardInterrupt:
336 except threadpool.NoResultsPending:
337 print "All results collected."
340 print FindbadPCURecordSync.query.count()
341 print FindbadPCURecord.query.count()
347 l_pcus = monitor.database.if_cached_else_refresh(1, config.refresh, "pculist", lambda : plc.GetPCUs())
350 fbsync = FindbadPCURecordSync.findby_or_create(plc_pcuid=0, if_new_set={'round' : global_round})
352 global_round = fbsync.round
355 # update global round number to force refreshes across all nodes
357 fbsync.round = global_round
359 if config.site is not None:
360 api = plc.getAuthAPI()
361 site = api.GetSites(config.site)
362 l_nodes = api.GetNodes(site[0]['node_ids'], ['pcu_ids'])
365 pcus += node['pcu_ids']
367 l_pcus = [pcu for pcu in sets.Set(pcus)]
368 elif config.pcuselect is not None:
369 n, pcus = pcu_select(config.pcuselect)
371 l_pcus = [pcu for pcu in sets.Set(pcus)]
373 elif config.nodelist == None and config.pcuid == None:
374 print "Calling API GetPCUs() : refresh(%s)" % config.refresh
375 l_pcus = [pcu['pcu_id'] for pcu in l_pcus]
376 elif config.nodelist is not None:
377 l_pcus = util.file.getListFromFile(config.nodelist)
378 l_pcus = [int(pcu) for pcu in l_pcus]
379 elif config.pcuid is not None:
380 l_pcus = [ config.pcuid ]
381 l_pcus = [int(pcu) for pcu in l_pcus]
383 checkAndRecordState(l_pcus, cohash)
388 if __name__ == '__main__':
390 logger = logging.getLogger("monitor")
391 logger.setLevel(logging.DEBUG)
392 fh = logging.FileHandler("monitor.log", mode = 'a')
393 fh.setLevel(logging.DEBUG)
394 formatter = logging.Formatter('%(asctime)s %(levelname)s %(message)s')
395 fh.setFormatter(formatter)
396 logger.addHandler(fh)
397 from monitor import parser as parsermodule
398 parser = parsermodule.getParser()
399 parser.set_defaults(nodelist=None,
404 dbname="findbadpcus",
408 parser.add_option("-f", "--nodelist", dest="nodelist", metavar="FILE",
409 help="Provide the input file for the node list")
410 parser.add_option("", "--site", dest="site", metavar="FILE",
411 help="Get all pcus associated with the given site's nodes")
412 parser.add_option("", "--pcuselect", dest="pcuselect", metavar="FILE",
413 help="Query string to apply to the findbad pcus")
414 parser.add_option("", "--pcuid", dest="pcuid", metavar="id",
415 help="Provide the id for a single pcu")
417 parser.add_option("", "--cachenodes", action="store_true",
418 help="Cache node lookup from PLC")
419 parser.add_option("", "--dbname", dest="dbname", metavar="FILE",
420 help="Specify the name of the database to which the information is saved")
421 parser.add_option("", "--refresh", action="store_true", dest="refresh",
422 help="Refresh the cached values")
423 parser.add_option("-i", "--increment", action="store_true", dest="increment",
424 help="Increment round number to force refresh or retry")
425 parser = parsermodule.getParser(['defaults'], parser)
426 config = parsermodule.parse_args(parser)
428 # NOTE: evidently, there is a bizarre interaction between iLO and ssh
429 # when LANG is set... Do not know why. Unsetting LANG, fixes the problem.
430 if 'LANG' in os.environ:
431 del os.environ['LANG']
434 except Exception, err:
435 traceback.print_exc()
436 print "Exception: %s" % err
437 print "Saving data... exitting."