X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=nodebad.py;h=dc866645eba29d39184df4544391e3b19bfea537;hb=675c7ae373a98e3c97f560a5b36978c36ee1b038;hp=c3aae39619d18335e985f1d62f2da2584d053ce5;hpb=6a452e8ece2ca8a47105c128eaebc38507bc76c5;p=monitor.git

diff --git a/nodebad.py b/nodebad.py
index c3aae39..dc86664 100755
--- a/nodebad.py
+++ b/nodebad.py
@@ -6,7 +6,7 @@ import string
 import time
 from datetime import datetime,timedelta
 
-from nodequery import verify,query_to_dict,node_select
+from monitor.query import verify,query_to_dict,node_select
 
 from monitor.common import *
 
@@ -40,43 +40,58 @@ def check_node_state(rec, node):
 	if rec.plc_node_stats:
 		print rec.plc_node_stats
 		boot_state = rec.plc_node_stats['boot_state']
+		run_level = rec.plc_node_stats['run_level']
 		last_contact = rec.plc_node_stats['last_contact']
+		node.plc_nodeid = rec.plc_node_stats['node_id']
 	else:
 		boot_state = "unknown"
 		last_contact = None
 
 	if boot_state == 'disable': boot_state = 'disabled'
-	if boot_state == 'diag': 	boot_state = 'diagnose'
+	if boot_state == 'diag' or boot_state == 'diagnose': boot_state = 'safeboot'
 
 	if len(rec.plc_node_stats['pcu_ids']) > 0:
 		node.haspcu = True
 	else:
 		node.haspcu = False
 
+	node.firewall = rec.firewall
+	node.plc_siteid = rec.plc_node_stats['site_id']
+
 	# NOTE: 'DOWN' and 'DEBUG'  are temporary states, so only need
 	# 			'translations' into the node.status state
 	#		'BOOT' is a permanent state, but we want it to have a bit of
 	#			hysteresis (less than 0.5 days)
-
 	#################################################################
 	# "Initialize" the findbad states into nodebad status if they are not already set
 
-	if node_state == 'DOWN' and ( node.status != 'offline' and node.status != 'down' ) and boot_state != 'disabled' :
-		print "changed status from %s to offline" % node.status
-		node.status = 'offline'
-		node.last_changed = datetime.now()
+	if node_state == 'DOWN':
+		if boot_state == 'disabled' and changed_lessthan(node.last_changed, 60) and \
+			node.status != 'disabled':
+			# NOTE: if changed less than 2 months, then we can allow this. 
+			# otherwise, apply 'down' status after greater than 2 months (below).
 
-	if node_state == 'DEBUG' and node.status != 'monitordebug' and \
-								 node.status != 'disabled' and \
-								 node.status != 'diagnose':
-		if boot_state != 'disabled' and boot_state != 'diagnose':
+			print "changed status from %s to %s" % (node.status, boot_state)
+			node.status = boot_state
+			node.last_changed = datetime.now()
 
-			print "changed status from %s to monitordebug" % (node.status)
-			node.status = "monitordebug"
+		if node.status not in ['offline', 'down', 'disabled']:
+			print "changed status from %s to offline" % node.status
+			node.status = 'offline'
 			node.last_changed = datetime.now()
+
+	if node_state == 'DEBUG':
+		if boot_state != 'disabled' and boot_state != 'safeboot':
+			print "changed status from %s to failboot" % (node.status)
+			current_status = "failboot"
 		else:
 			print "changed status from %s to %s" % (node.status, boot_state)
-			node.status = boot_state
+			current_status = boot_state
+
+		if current_status != node.status and \
+			current_status in ['failboot', 'disabled', 'safeboot']:
+
+			node.status = current_status
 			node.last_changed = datetime.now()
 
 	if node_state == 'BOOT' and node.status != 'online' and node.status != 'good':
@@ -88,8 +103,8 @@ def check_node_state(rec, node):
 	# Switch temporary hystersis states into their 'firm' states.
 	#	  online -> good		after half a day
 	#	  offline -> down		after two days
-	#	  monitordebug -> down  after 30 days
-	#	  diagnose -> monitordebug after 60 days
+	#	  failboot -> down  after 30 days
+	#	  safeboot -> failboot after 60 days
 	#	  disabled -> down		after 60 days
 
 	if node.status == 'online' and changed_greaterthan(node.last_changed, 0.5):
@@ -102,15 +117,15 @@ def check_node_state(rec, node):
 		node.status = 'down'
 		# NOTE: do not reset last_changed, or you lose how long it's been down.
 
-	if node.status == 'monitordebug' and changed_greaterthan(node.last_changed, 30):
+	if node.status == 'failboot' and changed_greaterthan(node.last_changed, 30):
 		print "changed status from %s to down" % node.status
 		node.status = 'down'
 		# NOTE: do not reset last_changed, or you lose how long it's been down.
 
-	if node.status == 'diagnose' and changed_greaterthan(node.last_changed, 60):
+	if node.status == 'safeboot' and changed_greaterthan(node.last_changed, 60):
 		print "changed status from %s to down" % node.status
-		# NOTE: change an admin mode back into monitordebug after two months.
-		node.status = 'monitordebug'
+		# NOTE: change an admin mode back into failboot after two months.
+		node.status = 'failboot'
 		node.last_changed = datetime.now()
 
 	# extreme cases of offline nodes