Using Cloud Services for Disaster Recovery:
Disaster Recovery:
For any operating organization/business, disaster is considered as a risk which may be in the form of a cyberattack to equipment failures to natural disasters. Since a disaster risk cannot be eliminated, it is important to have a disaster recovery procedure in place to continue business operations. Disaster Recovery Plan (DRP) and Business Continuity Plan (BCP) are integral parts of an overall risk management for any organization. A DRP is to be developed in conjunction with a BCP to ensure successful business operations. Maintaining continuity of business processes is the focused objective of a having a Disaster Recovery in place. An important aspect of preparing a DRP is to understand the types of risk faced by the organization. Some of the common disasters faced by IT industries includes the risks listed in figure1.
A Disaster Recovery service operates by replicating between two data centers: the primary data center and backup data center. The requirements for a Disaster Recovery include: Recovery Point Objective (RPO) and Recovery Time Objective (RTO). The RPO represents the most recent backup prior to a failure. The RPO time is based on the organization’s decision to backup data. RTO represents the time taken for a system to jump back online after a disaster occurs. This includes the time taken to detect the failure, prepare the servers in the backup site, initiate failed system and perform network reconfiguration.
Two Disaster Recovery models are in available to prevent failure in an organization: Traditional and Cloud-based service. A traditional service can be used as dedicated infrastructure or shared approach. In a dedicated approach, the infrastructure is solely assigned resulting in high costs and speed. In a shared approach which is also called as distributed approach, the infrastructure is assigned to multiple users resulting in low costs and low speed.
Disaster Recovery using Cloud:
On the contrary, a Cloud-based service offers both dedicated and shared model benefits with low cost and high speed. The figure2 compares the traditional and Cloud-DR models:
Cloud-DR reduces data synchronization between primary and backup sites, minimizes costs and increases independency between users and DR systems. The below table1 exhibits comparison between different Disaster Recovery categories.
In addition to features, a DR solution should have the potential to detect a risk in the event of a disaster and perform failure procedure to activate the backup site and revert control to primary site after the disaster has passed. A majority of DR procedures rely on manual detection and failover mechanisms. A Cloud-based system can simply determine the extent of a network failure by monitoring the primary site from cloud nodes distributed across different geographic regions and react accordingly. The below Figure3 illustrates an example of Cloud-DR in Replication and Failover modes. In a Cloud-based DR, the level of resources required depends on the Replication Mode or Failover Mode of the system. During normal operation the system is in Replication Mode and requires only low cost VM to act as DR Server. In the event of a Disaster, the system shifts to Failover Mode, which requires resources to support full system.
Benefits of using Cloud for Disaster Recovery:
Cloud-based DR offers multiple benefits when compared to other traditional DR approaches:
• The cloud's pay-as-you go model significantly brings down costs due to the diverse level of resources required before and during a disaster.
• Cloud resources can rapidly be included with fine granularity and have costs that scale easily without requiring substantial investments.
• The cloud oversees and maintains DR servers, bringing down IT expenses and lessening the impact of failures at the disaster/primary site.
• Cloud-based DR provides Disaster Recovery as a service (DRaaS) which is optimal for smaller and growing organizations that cannot afford equipment and real estate required for hot, warm and cold disaster recovery sites. The below figure4 illustrates the DRaaS model of Cloud.
• Using cloud makes it easier to retrieve files and application data.
• IT assets available in the cloud can be accessed from across the globe and the data can be restored.
• Using Cloud decreases the response time to a disaster and organizations can recover from the disaster within the cloud itself using automated tools.
• With the cloud, disaster recovery facilities can rapidly and effortlessly be moved to various parts of the world.
• Cloud offers Superior efficiency and reliability by using state-of-the-art facilities.
• The hardware, software and maintenance costs are taken care by the provider.
• Cloud-based backup is easier to manage and user friendly.
Additionally, virtualization provides salient features for disaster recovery:
• VM startup can be effortlessly atomized, bringing down recovery times after a disaster.
• Virtualization wipes out hardware dependencies, conceivably bringing down equipment necessities at the primary site.
• System agnostic state replication software can run outside of the VM, considering it as a black box.
Cons using Cloud as DR:
• Data cannot be accessed without internet availability.
• Depending on bandwidth availability, large amount of data cannot be transferred within a stipulated time.
• Full data recovery could take excessive time and result in stressing of local systems.
• In the event of a disaster in a vast area, multiple organizations may experience service interruption and have to be recovered by CSPs. In such events, related servers may not be able to handle excess load from various organizations.
• Privacy and Confidentiality is another major concern where the private data centers of enterprises would be failover by cloud environments.
Cloud-based Disaster Recovery Approach:
Cloud based Disaster Recovery plan can be setup using multiple strategies. One of the strategy involves the following key steps to be followed:
1. Evaluation of Data Protection:
The initial phase in implementing a DR, is to assess the organization's needs. The results of the assessment will be the main considerations that decide the framework and design needed to set up cloud-based data protection.
2. Choosing a Cloud Provider:
Once the data protection needs are determined, the subsequent step is to recognize cloud suppliers that can accommodate the necessities. For instance, a few suppliers will enable the replication of a VM, however won't host it. Similarly, there are suppliers that offer storage. If the goal is only to replicate data to cloud, subscribing to a storage-only plan is feasible.
3. Cost Consideration:
Once the cloud provider has been identified, cost estimation is to be considered. Every cloud service provider has a unique pricing model, but the total monthly cost comprises of:
• A monthly subscription fee
• The amount of Internet bandwidth used
• The amount of storage space consumed
• The number of VMs (or virtual processors)
The cost of the cloud service selected has a major impact on the cost of the cloud-based disaster recovery initiative. Online tools/calculators are available to estimate the costs.
4. Security Consideration:
Security consideration is equally important as cost consideration while choosing a cloud service provider. The optimal way to ensure security is to audit cloud providers. The cloud providers should be willing to furnish documentation in regard to their data centre protection strategies. The range of data backups and any supplemental backup is to be considered. Few of the elements to be considered while auditing Cloud providers are:
• Location
• Possible Events
• Power Grid/Communications Considerations and Contingencies
• Proximity to ‘prone-to-danger’ locations
• Vendor’s DR emergencies
5. Bandwidth Management
Bandwidth Management is critical for various reasons, including:
• Numerous cloud service providers charge for bandwidth utilization.
• Internet service provider may impose monthly usage caps or may charge for excess bandwidth usage.
• sufficient bandwidth is required to enable data to be backed up (or replicated) in a timely manner.
• It is to be ensured that cloud backups or replicated data don't expend high bandwidth that it effects other Internet usage.
6. Logistical Requirements:
In case of using cloud solely for storage capabilities, a minimal amount of logistical planning is required. If the organization wishes to perform failovers to the cloud, a number of considerations is to be taken into account which include: copying of data and Active Directory synchronization.
7. Virtual machine Replication:
An alternative approach for using Cloud-DR is to replicate VM’s to cloud. VM replication provides the following benefits:
• Point-in-time-image-based recovery
• Ability to mount a cloud-based copy of a VM and extract data
• Ability to redirect users to a cloud-based VM replica in the event of a disaster
Available Cloud Vendors in the market:
1. Bluelock – DRaaS can be tailored according to the organization’s needswhich include DRaaS Run, DRaaS Ready and DRaaS Restore. Tests have shown 100% success rate. Support team is available all-round the clock.
2. Codero’s DRaaS – offers 15-minute failover guarantee. Allows organizations to manage backup and DR storage costs by keeping recent backups on the local equipment while the rest is archived to the cloud.
3. NTT Communications – offers DRaaS by leveraging Microsoft Azure Site Recovery. Allows replication of enterprise workloads to an NTT Com data center or Microsoft Azure, which ensures real-time business continuity in the event of a disaster.
References:
1. Charlotte Brooks, Matthew Bedernjak, Igor Juran, John Merryman, (2002) “ Disaster Recovery Strategies”
2. https://searchdisasterrecovery.techtarget.com/feature/Implementing-cloud-based-disaster-recovery-Six-key-steps
3. https://solutionsreview.com/backup-disaster-recovery/5-disaster-recovery-as-a-service-vendors-to-watch-in-2018/
4. Mohammad Ali Khoshkholghi, Azizol Abdullah, Rohaya Latip, Shamala Subramaniam & Mohamed Othman, (2014) “Disaster Recovery in Cloud Computing: A Survey” – University Putra Malaysia, Selangor, Malaysia.
5. Mr.A.Srinivas, Y.Seetha Ramayya, B.Venkatesh, “A Study on Cloud Computing Disaster Recovery” – Coastal Institute of Technology & Management, Vizianagaram, India.
6. Somesh P. Badhel, Prof. Vikrant Chole, “A Review on Data Back-up Techniques for Cloud Computing”IJCSMC, Vol. 3, Issue. 12, December 2014, pg.538 – 542, G. H. Raisoni Academy of Engineering and Technology, Nagpur, India.
7. Timothy Wood, Emmanuel Cecchet, K.K Ramakrishnan, Prashant Shenoy, Jacobus van der Merwe and Arun Venkataramani, “Disaster Recovery as a Cloud Service: Economic Benefits & Deployment Challenges” – University of Massachusetts Amherst.
8. Yu Gu, Dongsheng Wang, and Chuanyi Liu, “DR-Cloud: Multi-Cloud Based Disaster Recovery Service” – Tsinghua Science And Technology.
...(download the rest of the essay above)