Certain clients with specific configurations and versions have had an issue where duplicate Load Balancer/HAProxy processes are causing a split-brain type of operation after reloads. Historically the fix has been stopping and starting the load balancer to clear it out, however, there is a patch available for the problem.
Important! Only apply this is you are running either Snapt_Leap_v11.x86_64-2.0.1 or Snapt_v13.x86_64-1.0.4.
You can check your version by running the following command:
cat /etc/os-release|grep PRETTY_NAME
You should see exactly this string: PRETTY_NAME=”openSUSE Leap 42.1 (x86_64)”
If you do you can now continue. Ensure you take a backup of your entire VM before running any patches!
Step 1: Stop the virtual Load Balancer
Ensure that you have stopped the load balancer completely in the UI, and if you are using redundancy monitored applications, remove it from there before stopping.
Step 2: Download and install the patch
wget https://shop.snapt.net/clientdata/snapt-ha-1.7.8-1.1.x86_64.rpm /etc/init.d/haproxy stop; service haproxy stop; killall -9 haproxy /bin/rpm -i --replacepkgs --replacefiles snapt-ha-1.7.8-1.1.x86_64.rpm
Step 3: Start the virtual Load Balancer
You may now start the load balancer again. Check everything is working correctly and test a reload.