changes for 3.0
[monitor.git] / findbadpcu.py
index 5f54235..114c48b 100755 (executable)
@@ -5,10 +5,14 @@ import sys
 import string
 import time
 import socket
+import util.file
+import plc
+import sets
 
     
 import signal
 import traceback
+from nodequery import pcu_select
 
 #old_handler = signal.getsignal(signal.SIGCHLD)
 
@@ -81,6 +85,8 @@ def get_pcu(pcuname):
                                        l_pcu = i
                except:
                        traceback.print_exc()
+                       from nodecommon import email_exception
+                       email_exception()
                        l_pcu = None
 
        plc_lock.release()
@@ -99,6 +105,8 @@ def get_nodes(node_ids):
                                        l_node.append(n)
                except:
                        traceback.print_exc()
+                       from nodecommon import email_exception
+                       email_exception()
                        l_node = None
 
        plc_lock.release()
@@ -156,6 +164,8 @@ def get_plc_site_values(site_id):
                                        break
                except:
                        traceback.print_exc()
+                       from nodecommon import email_exception
+                       email_exception()
                        values = None
 
        plc_lock.release()
@@ -194,6 +204,8 @@ def collectPingAndSSH(pcuname, cohash):
                except:
                        b_except = True
                        traceback.print_exc()
+                       from nodecommon import email_exception
+                       email_exception()
                        continue_probe = False
 
                if b_except or not continue_probe: return (None, None, None)
@@ -326,7 +338,7 @@ def checkAndRecordState(l_pcus, cohash):
        global count
        global_round = externalState['round']
 
-       tp = threadpool.ThreadPool(20)
+       tp = threadpool.ThreadPool(10)
 
        # CREATE all the work requests
        for pcuname in l_pcus:
@@ -380,11 +392,25 @@ def main():
                # update global round number to force refreshes across all nodes
                externalState['round'] += 1
 
-       if config.filename == None and config.pcuid == None:
+       if config.site is not None:
+               api = plc.getAuthAPI()
+               site = api.GetSites(config.site)
+               l_nodes = api.GetNodes(site[0]['node_ids'], ['pcu_ids'])
+               pcus = []
+               for node in l_nodes:
+                       pcus += node['pcu_ids']
+               # clear out dups.
+               l_pcus = [pcu for pcu in sets.Set(pcus)]
+       elif config.pcuselect is not None:
+               n, pcus = pcu_select(config.pcuselect)
+               # clear out dups.
+               l_pcus = [pcu for pcu in sets.Set(pcus)]
+
+       elif config.nodelist == None and config.pcuid == None:
                print "Calling API GetPCUs() : refresh(%s)" % config.refresh
                l_pcus  = [pcu['pcu_id'] for pcu in l_pcus]
-       elif config.filename is not None:
-               l_pcus = config.getListFromFile(config.filename)
+       elif config.nodelist is not None:
+               l_pcus = util.file.getListFromFile(config.nodelist)
                l_pcus = [int(pcu) for pcu in l_pcus]
        elif config.pcuid is not None:
                l_pcus = [ config.pcuid ] 
@@ -404,20 +430,26 @@ if __name__ == '__main__':
        formatter = logging.Formatter('%(asctime)s %(levelname)s %(message)s')
        fh.setFormatter(formatter)
        logger.addHandler(fh)
-       from config import config
-       from optparse import OptionParser
-       parser = OptionParser()
-       parser.set_defaults(filename=None, 
+       import parser as parsermodule
+       parser = parsermodule.getParser()
+       parser.set_defaults(nodelist=None, 
                                                increment=False, 
                                                pcuid=None,
+                                               pcuselect=None,
+                                               site=None,
                                                dbname="findbadpcus", 
                                                cachenodes=False,
                                                refresh=False,
                                                )
-       parser.add_option("-f", "--nodelist", dest="filename", metavar="FILE", 
+       parser.add_option("-f", "--nodelist", dest="nodelist", metavar="FILE", 
                                                help="Provide the input file for the node list")
+       parser.add_option("", "--site", dest="site", metavar="FILE", 
+                                               help="Get all pcus associated with the given site's nodes")
+       parser.add_option("", "--pcuselect", dest="pcuselect", metavar="FILE", 
+                                               help="Query string to apply to the findbad pcus")
        parser.add_option("", "--pcuid", dest="pcuid", metavar="id", 
                                                help="Provide the id for a single pcu")
+
        parser.add_option("", "--cachenodes", action="store_true",
                                                help="Cache node lookup from PLC")
        parser.add_option("", "--dbname", dest="dbname", metavar="FILE", 
@@ -426,8 +458,8 @@ if __name__ == '__main__':
                                                help="Refresh the cached values")
        parser.add_option("-i", "--increment", action="store_true", dest="increment", 
                                                help="Increment round number to force refresh or retry")
-       config = config(parser)
-       config.parse_args()
+       parser = parsermodule.getParser(['defaults'], parser)
+       config = parsermodule.parse_args(parser)
        try:
                # NOTE: evidently, there is a bizarre interaction between iLO and ssh
                # when LANG is set... Do not know why.  Unsetting LANG, fixes the problem.
@@ -437,6 +469,8 @@ if __name__ == '__main__':
                time.sleep(1)
        except Exception, err:
                traceback.print_exc()
+               from nodecommon import email_exception
+               email_exception()
                print "Exception: %s" % err
                print "Saving data... exitting."
                database.dbDump(config.dbname, externalState)