|
||
In this age of information technology, companies tend to plan for the worst and hope for the best. With dependency on data on a rise, it would be hard to imagine what our business and operations would look like in the event of a data loss disaster.
Customers are demanding business solutions 24X7, and businesses have to meet this demand in order to sustain their position in the market. Therefore, business downtime can spell disaster. Due to this reason, companies are spending huge amounts of money on disaster recovery planning.
What is Disaster Recovery Planning?
Disaster recovery planning, as the name suggests, is a comprehensive plan that highlights how the business can resume its normal course of business in the event of a disaster. Data disasters can strike at any time and they are most likely to strike your business when you are least prepared. A critical role of disaster recovery planning is to showcase how a business can regain access to its important data systems and other resources so that the business can run as usual.
Objectives of Disaster Recovery Planning
The main objective of disaster recovery planning is to ensure that the business can function normally amidst the disaster crisis. However, there are other objectives of disaster recovery planning such as the safety and security of the human resources of a company.
In addition, the plan should mention what should be the procedures to be followed on a timely basis to ensure that the impact of the disaster can be minimised. For example, a disaster recovery plan should encompass details about data systems security and data backup techniques to ensure that the data is regularly backed up. Therefore, one of the major objectives of disaster recovery planning is to limit data loss and aid data recovery in the time of crisis.
Natural and Human-Induced Disasters that can Result in Data Loss
Natural disasters in the recent past have been responsible for huge business losses. Hurricane Katrina and the recent forest fires in California are responsible for of large-scale business losses resulting from natural disasters. There are other forms of natural disasters such as earthquakes, building collapse, cyclones, volcano eruptions, and floods that can also cause severe data loss disasters.
The most horrific form of human-induced disaster has been the 9/11 terror attacks. Such form of terror attacks can result in large-scale data destruction and only off-site backups can save the day. There are also cases of widespread virus attacks that affect millions of computer data systems and have the power to bring the whole business to a halt. Such viruses can result in huge business downtime.
Data Loss Prevention Measures
While there is little you can do to avoid the data loss incident completely, you can limit your business's exposure to data loss by adopting certain preventive measures such as remote backup, micro-filming of sensitive documents, creation of storage area networks, installation of surge protectors, fire protection measures, etc.
•Remote Backup
Companies are increasingly opting for a remote backup of their critical data to ensure that even if a disaster occurs at a primary business site, the data remains protected.
•Micro Filming of Sensitive Documents
Micro filming of sensitive documents ensures that the data stays protected when a data loss disaster occurs. These records should be stored offsite.
•Creation of Storage Area Networks (SAN)
SAN allows data storage devices to exist on a separate network and communicate with each other directly using a very fast media. Server systems are connected to a LAN as well as a SAN to enable the users to gain access to these data storage devices.
•Installation of surge protectors
Sudden electric fluctuations can cause hard drive failures resulting in data loss. Therefore, your business should protect your valuable data from these sudden power surges via installations of surge protectors.
What are SANs? How Can They Help?
SANs are thought to be an ideal way to backup as well as recover your data. SAN devices are physically apart from the main data storage site but are still connected to the network. Therefore, SAN devices provide a perfect separate network to exclusively store the data and make the data available immediately after the disaster. SANs also help during the data recovery process.
What are Hot Sites, Cold Sites, and Warm Sites?
•Cold sites are sites where a business can relocate after the disaster. A cold site does not have the hardware required to kick-start the business after the disaster nor does it encompass any backed up data. This makes cold site the cheapest backup site for a business. However, the business downtime will be more if a business uses a cold site.
•Warm Sites are sites where a business can relocate after the disaster strikes. A warm site has the requisite hardware equipment to kick-start the business. However, it does not have any backed up data.
•Hot Sites are the most expensive of all the three options. A hot site is an exact replica of the original business site with the requisite computer systems as well as backed up data. Business can resume operations instantly if they own a hot site. However, these are very expensive to maintain.
A disaster is, by its definition, a low odds probability with a high cost if it happens. An astonishing number of businesses, caught up in the day-to-day running of the enterprise, don't take adequate steps for disaster prevention or have adequate recovery procedures in place. Disasters can be natural (earthquakes, hurricanes, tornadoes) or manmade (vandalism by a former employee, terrorist threats). Either way, your organization's data centre is its brain, and just like a person who has taken a severe brain injury, most organisations that suffer a severe data centre disruption don't recover.
Here are some generally agreed upon best practices. Whilst they are not a sovereign remedy against disaster, they are the measures you should take to ensure continuity.
First, have a plan. Understand what risks your location presents – while hurricanes aren't likely, earthquakes are possible in the UK. Your plan should include, at the very least, who you contact in the event of a service disruption to your power supply or your connectivity or both.
Second, your plan should include provisions for the people who work in your data centre. Will they need facilities to sleep in when the disaster strikes? Or a place where their family can be safe? Your data centre has people, and they are the most important asset you have. Similarly, if you have a mission critical data centre, look at ways to generate your own power; many in the US have on-site diesel storage tanks and generation facilities, so that when the power runs out, before the batteries run out, they can generate onsite storage.
Your next most important asset is the data, and the mantra of data preservation is triage, triage, triage, and backup, backup, backup.
Triage means making a list of what data needs to be backed up and what data doesn't. For example, for your servers, build a mirror that you can restore from that contains the bare operating system, the required applications and configuration files – and then nothing else. That is your restore point; from there the only data you need to back up on your production servers are anything that differs from the default installation.
This can save you tremendous amounts of labour (and confusion) when backing up data; far too many organisations just blindly copy the contents of all hard drives rather than make staged backups.
Your backup plan should include both on site and off site backup provisions. You should make backups on servers at your facility regularly, and should mirror them on an off-site facility as well, so that something that damages your facility doesn't destroy all your data.
The most critical part of a backup plan is doing test restores. You absolutely need to schedule a test restore from backups at least once per month if not more regularly; doing backups without test restorations is a sure fire way to discover, at the wrong time, that your restoration technique is inadequate at the worst possible time. Your users will complain about having a day interrupted for a test restore every month, but it is essential to making this process work.