Blame SOURCES/bz1760213-fence_compute-disable-service-after-force-down.patch

6147c0
From 099758a41bbb153c4a13a89de57cdcb72e1f1ea7 Mon Sep 17 00:00:00 2001
6147c0
From: Michele Baldessari <michele@acksyn.org>
6147c0
Date: Fri, 11 Oct 2019 10:39:53 +0200
6147c0
Subject: [PATCH] fence_compute: Invert the force-down/service disable order
6147c0
6147c0
In OpenStack Train we first observed that IHA was not working via
6147c0
https://bugzilla.redhat.com/show_bug.cgi?id=1760213
6147c0
6147c0
The reason for this is that nova has made the disabling of the compute
6147c0
service depend on the compute node being up via:
6147c0
https://review.opendev.org/#/c/654596/
6147c0
6147c0
By first calling force-down, the subsequence service-disable API
6147c0
call won't wait for the reachability of the compute node any
6147c0
longer and the whole operation has the same outcome.
6147c0
6147c0
Tested this on an OSP Train environment and we correctly
6147c0
got Instance HA working again and we observed the VMs being
6147c0
restarted on the available compute nodes.
6147c0
6147c0
Co-Authored-By: Luca Miccini <lmiccini@redhat.com>
6147c0
---
6147c0
 agents/compute/fence_compute.py | 2 +-
6147c0
 1 file changed, 1 insertion(+), 1 deletion(-)
6147c0
6147c0
diff --git a/agents/compute/fence_compute.py b/agents/compute/fence_compute.py
6147c0
index a94bdc46..c08a9dbe 100644
6147c0
--- a/agents/compute/fence_compute.py
6147c0
+++ b/agents/compute/fence_compute.py
6147c0
@@ -117,11 +117,11 @@ def set_power_status_off(connection, options):
6147c0
 	if status in [ "off" ]:
6147c0
 		return
6147c0
 
6147c0
-	connection.services.disable(options["--plug"], 'nova-compute')
6147c0
 	try:
6147c0
 		# Until 2.53
6147c0
 		connection.services.force_down(
6147c0
 			options["--plug"], "nova-compute", force_down=True)
6147c0
+		connection.services.disable(options["--plug"], 'nova-compute')
6147c0
 	except Exception as e:
6147c0
 		# Something went wrong when we tried to force the host down.
6147c0
 		# That could come from either an incompatible API version