From baa9fb8c4b2b993b4d5886dc014c92667727d369 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?C=C3=A9dric=20Jeanneret?= <cjeanner@redhat.com>
Date: Tue, 16 Apr 2019 13:34:03 +0200
Subject: [PATCH] Correct ovn-dbs health check

We can't really check the socket presence in containers:
- ovn doesn't really listen to them
- playing with lsof and ss doesn't help, since there are some
  issues with inodes and overlays

The new healthcheck ensures the service is properly running, and
will fail if ovn-northd has an issue.

Please note: the current STDERR has some output, this is due to
some packaging issue being worked on right now.

Change-Id: I645e18cf198a948479083df94b5d373ed92f2aae
Closes-Bug: #1823882
---
 healthcheck/ovn-dbs | 11 +----------
 1 file changed, 1 insertion(+), 10 deletions(-)

diff --git a/healthcheck/ovn-dbs b/healthcheck/ovn-dbs
index 46b78d84d..4a0b4de62 100755
--- a/healthcheck/ovn-dbs
+++ b/healthcheck/ovn-dbs
@@ -1,13 +1,4 @@
 #!/bin/bash
 
-. ${HEALTHCHECK_SCRIPTS:-/usr/share/openstack-tripleo-common/healthcheck}/common.sh
-
 process='ovn-northd'
-sockets="${@:-/run/openvswitch/ovnnb_db.sock /run/openvswitch/ovnsb_db.sock}"
-
-for sock in sockets; do
-    if ! healthcheck_socket $process $sock; then
-        echo "There is no $process process connected to socket $sock running in the container"
-        exit 1
-    fi
-done
+/usr/bin/ovs-appctl -t ${process} version 1>/dev/null