fbpx
The Stack Archive

What should I really be looking for in a disaster recovery solution?

Wed 23 Sep 2015

Office hit by natural disaster

julestaplin22Jules Taplin has a long experience of dealing with users over their IT disaster recovery provision. Here he suggests important elements to consider when looking at DR solutions…

For smaller companies, without the technical expertise on the subject, sourcing a suitable disaster recovery solution can be a mind-blowing task. It is no longer about ticking boxes to appease the board and compliance departments, it’s about finding the solution that will one day save your business and get you out of a lot of trouble.

When sourcing DR for your business it’s about prioritising what’s important for you, and here are a few elements to consider:

Managing your DR in-house or leaving it to the experts

If you have the resource you may be tempted to manage your own DR – buy a backup or replication product and off you go. You’ll need to ensure that you’re testing the replications regularly, and every so often running a full DR test to application level. When a disaster strikes it will be down to you to recover your own systems, and the success of your recovery will lie in the frequency of your testing.

If you opt for a fully managed service (whether this be a company managing a third party replication product for you or a company working on their own technology) then you can leave all this to the experts. They will ensure your DR is set-up properly, working, tested regularly and, most importantly, in a recovery, will be there to perform your recovery for you.

It’s easy to forget that during an IT failure it’s often much better for a third party expert (who will not be stressed by the impact of the failure) to manage the recovery while an internal IT department manages communications with the business and gets the employees back to productivity as soon as possible.

Coping with risk

This may sound like a silly question but ask yourself – if my recovery doesn’t go smoothly and I am without my IT systems for a few days, can my business cope?

It’s important to identify your appetite for risk. If you have a very low appetite for risk, then the only thing you can do is opt for a solution which is highly tested. You can now get solutions which test every day to an application level without disrupting your business. This will never happen if you run your DR in-house, as it is only cost effective through automation. If you have a higher appetite for risk then you may feel that running a yearly DR test to application level is sufficient, with more regular tests for data corruption which can be done in-house or by outsourcing through most providers.

Recovery times

32% of companies have a recovery time of between 4 and 24 hours [PDF]. With many DR companies offering a 4 hour RTO as standard, it’s not surprising that this has become industry standard. But have a think about what you need for your business – would you like to have critical services back quicker? Your budget could be proportioned accordingly so that critical services such as email, web servers and operations systems have a recovery time within an hour, yet less critical systems such as development servers have a recovery time of 4 days. You can now effectively pick and choose recovery times for individual servers, tailoring your disaster recovery service to meet your exacting needs.

Balancing data loss and recovery time

Whilst instant recovery time and zero data loss is the holy grail of DR, this is a very expensive solution which requires a doubling of your IT budget, as you need to duplicate your entire live system and invest in the maintenance of your standby, as you would your live system. This is cost prohibitive for most businesses, so in reality you may need to compromise on either recovery time (RTO) or recovery point (RPO).

If you are handling lots of transactional data, for example in the financial sector, then data loss is likely to be more critical. With 15-minute replications you can have RPO’s as low as 15 minutes, but it may take a little while for you to get your IT systems back, as you’ll be recovering from your most recent replication.

If however, you would rather get your IT systems back quickly and can afford to lose a couple of hours of data, then recovery time should be your priority, and you could be back working within just a few minutes.

Once you’ve decided on your priorities you can then start looking for the supplier that suits your needs. With such a range of options out there, just make sure you stick to your priorities. Opting for a range of solutions and suppliers is becoming a much more popular model as it enables companies to have a solution tailored precisely to their needs.

Jules Taplin is Technical Director with Plan B Disaster Recovery

Tags:

disaster recovery feature
Send us a correction about this article Send us a news tip