Microsoft Azure: The Simplest Load Balancing Cluster with Failover on Windows and Linux

Azure template for a quick deployment of a farm cluster in Microsoft Azure on Windows or Linux

The Evidian SafeKit farm cluster has been validated by Microsoft Azure in quickstart templates.

To deploy the Evidian SafeKit load balancing cluster with failover in Microsoft Azure, just click on the following button which deploys everything:

  Deploy to Microsoft Azure a Farm Cluster (Windows or Linux)  >

Go to the Template Guide tab for more information

Automatic installation in Microsoft Azure of a load balancing cluster with failover (Windows or Linux)

• Automatic deployment
• Configure deployment
• After deployment
• Video of the deployment
• Access VMs in Microsoft Azure
• Deployed resources

Automatic deployment of the Microsoft Azure template for a farm cluster

To deploy the Evidian SafeKit load balancing cluster with failover in Microsoft Azure, just click on the following button which deploys everything:

Deploy to Azure a Farm Cluster (Windows or Linux)  >

Configure the Microsoft Azure template for a farm cluster

After the click:

• in "Resource group", click on "Create new" and set a name
• choose the geographical "Location" where the cluster will be deployed
• choose the "OS" Windows or Linux
• choose the number of "Cluster Nodes"
• choose an "Admin User" name (not Administrator, not root)
• choose an "Admin Password". Passwords must be between 12 and 72 characters and have 3 of the following: 1 lower case, 1 upper case, 1 number, and 1 special character.
• click on "I agree..." and then on "Purchase" (no fee on SafeKit free trial, only on Microsoft Azure infrastructure)
• wait the end of deployment of the load balancing and failover cluster

After deployment

After deployment, click on 'Microsoft.Template' (previous image), then go to the output panel and

• visit the credential URL to install the client and CA certificates in your web browser. Force the load of the unsafe page. Put as user 'CA_admin' and the password you enter during the template configuration. Be careful, put the second certificate in the 'Trusted Root Certification Authority' store
• after certificates installation, start the web console of the cluster
• test the load balanced virtual IP address with the test URL in the output. A load balancing rule has been set for external port 9453, internal port 9453. A mosaic of server names is displayed according the server answering to the TCP session

Video of the Microsoft Azure farm template deployment

Accessing the VMs through SSH (Linux) or remote desktop (Windows)

If you want to connect to Virtual Machines through SSH (Linux) or remote desktop (Windows), you can use the SafeKit web console to know IP addresses or DNS names of VMs (next images). Use the user/password entered during the template configuration for accessing the VMs.

Deployed resources

In term of VMs, this template deploys:

• from 2 to 4 VMs (Windows or Linux)
• each VM has a public IP address
• the SafeKit free trial is installed in all VMs
• a SafeKit farm module is configured in all VMs

In term of load balancer, this template deploys:

• a public load balancer
• a public IP is associated with the public load balancer and plays the role of the virtual IP
• alls VMs are in the backend pool of the load balancer
• a health probe checks the farm module state on all VMs
• a load balancing rule for external port 9453 / internal port 9453 is set to test the load balanced virtual IP

For a manual installation, go to the Manual Installation tab

Manual installation on existing VMs in Microsoft Azure of a load balancing cluster with failover (Windows or Linux)

• Configuration of the Microsoft Azure load balancer
• Configuration of the Microsoft Azure network security
• Package installation on Windows
• Package installation on Linux
• Configuration of SafeKit
• Tests

Configuration of the Microsoft Azure load balancer

The load balancer must be configured to periodically send health packets to virtual machines. For that, SafeKit provides a health probe which runs inside the virtual machines and which

• returns OK when the farm module state is UP (green)
• returns NOT FOUND in all other states

You must configure the Microsoft Azure load balancer with:

• HTTP protocol
• port 9010, the SafeKit web server port
• URL /var/modules/farm/ready.txt (if farm is the module name that you will deploy later)

For more information, see the configuration of the Microsoft Azure load balancer.

Configuration of the Microsoft Azure network security

The network security must be configured to enable communications for the following protocols and ports:

• UDP - 4800 for the safeadmin service (between SafeKit nodes)
• TCP – 9010 for the load-balancer health probe and for the SafeKit web console running the http mode
• TCP – 9001 to configure the https mode for the console
• TCP – 9453 for the SafeKit web console running in the https mode

Package installation on Windows

On both Windows servers

• Install the free version of SafeKit for Cloud (click here) on 2 Windows nodes
• The module farm.safe is delivered inside the package.
• To open firewall, start a command line as administrator, goto C:\safekit\private\bin and type  .\firewallcfg.cmd add on both nodes

Package installation on Linux

On both Linux servers

• Install the free version of SafeKit for Cloud (click here) on 2 Linux nodes
• After the download of safekit_xx.bin package, execute it to extract the rpm and the safekitinstall script and then execute the safekitinstall script
• Answer yes to firewall automatic configuration
• The module farm.safe is delivered inside the package.

Configuration of SafeKit

The configuration is presented with the web console connected to 2 Windows servers but it is the same thing with 2 Linux servers.

Important: all the configuration must be done from a single browser.

It is recommended to configure the web console in the https mode by connecting to https://<IP address of 1 VM>:9453 (next image). In this case, you must configure before the https mode by using the wizard described in the User's Guide: see "11.1 HTTPS Quick Configuration with the Configuration Wizard".

Or you can use the web console in the http mode by connecting to http://<IP address of 1 VM>:9010 (next image).

Note that you can also make a configuration with DNS names, especially if the IP addresses are not static.

Enter IP address of the first node and click on Confirm (next image)

Click on New node and enter IP address of the second node (next image)

Click on the red floppy disk to save the configuration (previous image)

In the Configuration tab, click on farm.safe then enter farm as the module name and Confirm (next images with farm instead of xxx)

Click on Validate (next image)

Do not configure a virtual IP address (next image) because this configuration is already made in the Microsoft Azure load balancer. This section is useful for on-premise configuration only.

If a process is defined in the Process Checker section (next image), it will be monitored with the action restart in case of failure. The services will be stopped an restarted locally on the local server if this process disappears from the list of running processes. After 3 unsuccessful local restarts, the module is stopped on the local server. As a consequence, the health probe answers NOT FOUND to the Microsoft Azure load balancer and the load balancing is reconfigured to load balance the traffic on the remaining servers of the farm.

start_both and stop_both (next image) contain the start and the stop of services.

Click on Validate (previous image)

Click on Configure (previous image)

Check the success green message on both servers and click on Next (previous image)

Start the cluster on both nodes (previous image). Check that the status becomes UP (green) - UP (green) (next image).

The cluster is operational with services running on both UP nodes (previous image).

Be careful, components which are clients of the services must be configured with the virtual IP address. The configuration can be made with a DNS name (if a DNS name has been created and associated with the virtual IP address).

Tests

Check with Windows Microsoft Management Console (MMC) or with Linux command lines that the services are started on both UP nodes. Put  services with Boot Startup Type = Manual (SafeKit controls start of services).

Stop one UP node by scrolling down the menu of the node and by clicking on Stop. Check that the load balancing is reconfigured with only the other node taking all TCP connections. And check that the services are stopped on the STOP node with Windows Microsoft Management Console (MMC) or with Linux command lines.

More information on tests in the User's Guide

Automatic start of the module at boot

Configure boot start (next image on the right side) configures the automatic boot of the module when the server boots. Do this configuration on both nodes once the load balancing and failover solution is correctly running.

Note that for synchronizing SafeKit at boot and at shutdown on Windows, we assume that a command line has been run on both nodes during installation as administrator: .\addStartupShutdown.cmd in C:\safekit\private\bin (otherwise dot it now).

For reading the SafeKit logs, go to the Troubleshooting tab

For editing userconfig.xml, start_both and stop_both, go to the Advanced Configuration tab

Internal files of a SafeKit / Microsoft Azure load balancing cluster with failover

Internal files of the Windows farm.safe module

<!DOCTYPE safe>
<safe>
<service mode="farm" maxloop="3" loop_interval="24">
  <!-- Farm topology configuration -->
  <!-- Names or IP addresses on the default network are set during initialization in the console -->
  <farm>
    <lan name="default" />
  </farm>
  <!-- Software Error Detection Configuration -->
  <!-- Replace
       * PROCESS_NAME by the name of the process to monitor
  -->
  <errd polltimer="10">
    <proc name="PROCESS_NAME" atleast="1" action="restart" class="both" />
  </errd>
  <!-- User scripts activation -->
  <user nicestoptimeout="300" forcestoptimeout="300" logging="userlog" />
</service>
</safe>

@echo off

rem Script called on all servers for starting applications

rem For logging into SafeKit log use:
rem "%SAFE%\safekit" printi | printe "message"

rem stdout goes into Application log
echo "Running start_both %*" 

set res=0

rem Fill with your services start call

set res=%errorlevel%
if %res% == 0 goto end

:stop
set res=%errorlevel%
"%SAFE%\safekit" printe "start_both failed"

rem uncomment to stop SafeKit when critical
rem "%SAFE%\safekit" stop -i "start_both"

:end

@echo off

rem Script called on all servers for stopping application

rem For logging into SafeKit log use:
rem "%SAFE%\safekit" printi | printe "message"

rem ----------------------------------------------------------
rem
rem 2 stop modes:
rem
rem - graceful stop
rem   call standard application stop with net stop
rem
rem - force stop (%1=force)
rem   kill application's processes
rem
rem ----------------------------------------------------------

rem stdout goes into Application log
echo "Running stop_both %*" 

set res=0

rem default: no action on forcestop
if "%1" == "force" goto end

rem Fill with your services stop call

rem If necessary, uncomment to wait for the real stop of services
rem "%SAFEBIN%\sleep" 10

if %res% == 0 goto end

"%SAFE%\safekit" printe "stop_both failed"

:end

Internal files of the Linux farm.safe module

<!DOCTYPE safe>
<safe>
<service mode="farm" maxloop="3" loop_interval="24">
  <!-- Farm topology configuration for the membership protocol -->
  <!-- Names or IP addresses on the default network are set during initialization in the console -->
  <farm>
    <lan name="default" />
  </farm>
  <!-- Software Error Detection Configuration -->
  <!-- Replace
       * PROCESS_NAME by the name of the process to monitor
  -->
  <errd polltimer="10">
    <proc name="PROCESS_NAME" atleast="1" action="restart" class="both" />
  </errd>
  <!-- User scripts activation -->
  <user nicestoptimeout="300" forcestoptimeout="300" logging="userlog" />
</service>
</safe>

#!/bin/sh
# Script called on the primary server for starting application

# For logging into SafeKit log use:
# $SAFE/safekit printi | printe "message" 

# stdout goes into Application log
echo "Running start_both $*" 

res=0

# Fill with your application start call

if [ $res -ne 0 ] ; then
  $SAFE/safekit printe "start_both failed"

  # uncomment to stop SafeKit when critical
  # $SAFE/safekit stop -i "start_both"
fi

#!/bin/sh
# Script called on the primary server for stopping application

# For logging into SafeKit log use:
# $SAFE/safekit printi | printe "message" 

#----------------------------------------------------------
#
# 2 stop modes:
#
# - graceful stop
#   call standard application stop
#
# - force stop ($1=force)
#   kill application's processes
#
#----------------------------------------------------------

# stdout goes into Application log
echo "Running stop_both $*" 

res=0

# default: no action on forcestop
[ "$1" = "force" ] && exit 0

# Fill with your application stop call

[ $res -ne 0 ] && $SAFE/safekit printe "stop_both failed"