fence-agents/0009-fence_compute-Fix-disabling-force_down-on-node-when-.patch

From 832011f2287f1a9f59da727312e38fb9622dc809 Mon Sep 17 00:00:00 2001
From: Vincent Untz <vuntz@suse.com>
Date: Wed, 3 Feb 2016 13:45:52 +0100
Subject: [PATCH 9/9] fence_compute: Fix disabling force_down on node when
 action is on

When the action is on, the goal is to disable force_down for the
nova-compute service on the node.

However, we were only doing that if the nova-compute service was up;
which is impossible if it's forced to be down... So just always disable
force_down, and then, if it's up, do more things.
---
 fence/agents/compute/fence_compute.py | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)

diff --git a/fence/agents/compute/fence_compute.py b/fence/agents/compute/fence_compute.py
index e3d5061..98ad177 100644
--- a/fence/agents/compute/fence_compute.py
+++ b/fence/agents/compute/fence_compute.py
@@ -160,23 +160,23 @@ def set_power_status(_, options):
 		return
 
 	if options["--action"] == "on":
+		try:
+			# Forcing the host back up
+			nova.services.force_down(
+				options["--plug"], "nova-compute", force_down=False)
+		except Exception as e:
+			# In theory, if force_down=False fails, that's for the exact
+			# same possible reasons that below with force_down=True
+			# eg. either an incompatible version or an old client.
+			# Since it's about forcing back to a default value, there is
+			# no real worries to just consider it's still okay even if the
+			# command failed
+			logging.info("Exception from attempt to force "
+					  "host back up via nova API: "
+					  "%s: %s" % (e.__class__.__name__, e))
 		if get_power_status(_, options) == "on":
 			# Forcing the service back up in case it was disabled
 			nova.services.enable(options["--plug"], 'nova-compute')
-			try:
-				# Forcing the host back up
-				nova.services.force_down(
-					options["--plug"], "nova-compute", force_down=False)
-			except Exception as e:
-				# In theory, if force_down=False fails, that's for the exact
-				# same possible reasons that below with force_down=True
-				# eg. either an incompatible version or an old client.
-				# Since it's about forcing back to a default value, there is
-				# no real worries to just consider it's still okay even if the
-				# command failed
-				logging.info("Exception from attempt to force "
-					      "host back up via nova API: "
-					      "%s: %s" % (e.__class__.__name__, e))
 		else:
 			# Pretend we're 'on' so that the fencing library doesn't loop forever waiting for the node to boot
 			override_status = "on"
-- 
2.8.3