[pve-devel] [PATCH ha-manager 6/6] read_resources_config: ensure we got uptodate data

[Thomas Lamprecht]

minimizes the chance that we get outdated data here thus removing all
services as they are "without a node".

Signed-off-by: Thomas Lamprecht <t.lamprecht at proxmox.com>
---

This is normally only a problem if pmxcfs restarts, so I asked myself
if it we should remember the last restart from pmxcfs, and trigger an
cfs_update there if we enter in a method where it could matter and
the a restart was detected (i.e., the restart time is newer).

 src/PVE/HA/Config.pm | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/PVE/HA/Config.pm b/src/PVE/HA/Config.pm
index bf37b04..730e6a5 100644
--- a/src/PVE/HA/Config.pm
+++ b/src/PVE/HA/Config.pm
@@ -87,6 +87,7 @@ sub read_resources_config {
 # checks if resource exists and sets defaults for unset values
 sub read_and_check_resources_config {
 
+    PVE::Cluster::cfs_update();
     my $res = cfs_read_file($ha_resources_config);
 
     my $vmlist = PVE::Cluster::get_vmlist();
-- 
2.11.0