Customer Challenges

The customer, a manufacturing firm had a cloud hosted application.

• During heavy data traffic, the application was unable to scale dynamically to support the traffic
• Since automated backup was not available, daily manual back up was getting very time consuming as the application data was growing in volume.
• Any network or hardware failure, resulted in unexpected interruption was unavailable. The downtime in the application incurred huge monitory losses.

Key Technologies

• Ubuntu Server
• Azure Load balancer
• NSG
• Azure Active Directory
• Clustering and Resource Monitoring
• Continuous Data Replication for Failover
• Open SUSE

Software Used

• Server application ( Node JS / DotNET/ Php )
• Corosync, Pacemaker ( Clustering )
• Hawk Console ( Resource Monitoring )
• Hana Database

Solution Highlights

• Using Azure Load balancer, VMs were configured for high availability
• Clustering for both the VMs at OS level using pacemaker and corosync
• Automated Data replication from the primary server to secondary server to ensure data consistency
• Monitoring cluster nodes using Hawk
• Configuring health probes and backend pools for scaling up the application

Implementation Steps

• Firstly, a resource group on Azure was created. Then, a VM was added to that specific resource group to add all the created resources
• Two Opensuse VMs were created in availability set. A load balancer was configured on the backend to Opensuse.
• Opensuse VM packages were updated, and then the required packages were installed for clustering. By configuring both the VMS, clustering was enabled between two the VMs at OS level.
• HAWK Console was used to monitor the created resources.
• HANA database was installed in both VMs and SAP Hana studio was used for clustering and replication was performed from primary to secondary database.
• The resources to be monitored were then configured from Azure and across the OS level.
• After this, topology of master and slave was configured.
• An Active Directory application was created and configured to the created VMs which allowed us to monitor the application from azure.
• After this level of configuring clustering replication,
• Successful failover and failback tests were performed
• The data that was replicated was verified

Benefits

• Biased routing of traffic – effective routing of the traffic during peak load by distributing the load among the 2 servers
• Automated data backup – with the data being backed up automatically, disaster recovery management is in place
• High availability – for seamless back up and failover
• Cluster Monitoring/ Management – view and analyze the logs to assess the performance of the clusters
• Automatic reconfiguration & Dynamically Scalable
• Reliable protection for mission critical data since high availability even during down time
• Robust management across all platforms.