Colin Fettes from Azamour Solutions takes a look at alternative ways to handle Disaster Recovery (DR) in this year’s Arcati Mainframe Yearbook. His article focuses on recovery services, and hope that these thoughts help to clarify your own ideas and create some useful assistance around DR.
Perhaps the most interesting current trend is the level of appetite within the mainframe community for exploring new and more flexible approaches to disaster recovery. The reasons for this are the subject of much speculation and there are probably many – perhaps DR has become a viable target for cost reduction in the eyes of many organizations, maybe the same from the main providers (and changes to the levels of service offered), in the face of the ever increasing more general problems facing data centres, eg the provision of power, their physical location and staff access.
Let’s consider some of the factors you might want to think about when looking at DR provision with fresh eyes – there are a large number of these and far more implications than that, so unfortunately this won’t be an exhaustive list!
Location – this is probably the first thing that springs to mind when considering the task at hand. In ‘the old days’ this decision would have been based on the recovery centre being far enough away from the production site to avoid the plane crash, but a reasonable distance for staff to travel.
It was very much normal for the decision on a supplier to be driven by the location of their data centre and as mainframe DR was relatively ‘simple’ and that was that. How times have changed, it’s a bit more complex than that now, and some more considerations might be:
- Ability to travel (staff access)
- Risks eg proximity to airports, flooding, etc
- Access to specific networking capability eg MPLS
- Access to redundant power or communications.
Security – yes that’s normally the second concern, and another area where modern thinking has changed perceptions. Where once customers were seeking the most secure facility that they could possibly afford to use, short of armed guards and a self-destruct facility, more recently there has been more of a focus on facilities that are more people-friendly, and easier and quicker for staff to actually get in to, while retaining the levels of security that customers require – the result is the rise of the tier-3 data centre as a recovery centre.
In some sense there’s very little difference between what different suppliers and data centres can offer to the customer, and it’s often about the cost of the capability that is required rather than if that capability is on the table.
Some items worthy of note are:
- Would adding capability such as data replication provide better value to the business?
- Are truly redundant communications available, or access to the provider I want to use?
- Do I require relatively new technology equipment (this can be very important to mainframe users)?
- Are seats available within the data centre, is that the best value, and are they so oversubscribed that I would never actually get to use them?
- Am I reliant on my internal staff for testing / recovery or can the service provider offer assistance, eg my DR plan may require an ‘experienced sysprog’ to implement?
- Are there suitable facilities for people, eg local accommodations and places to eat?
- Are the relevant services available to assist us at test and invocation time?
For many users this is now becoming the key area and, in my experience, often the reason for issues with an existing DR provider.
Some of the questions we’ve been asked by potential customers include:
- Can I have the number of tests annually that I would like?
- Can I make minor changes to equipment quickly?
- Can communications services be modified, up / downgraded as my needs change?
- Are appropriate configuration tools available or can I use my own?
- Can I take the option of dedicated storage?
- Can we use our laptops or other equipment in the data centre?
- Can I add functionality that I may need as I grow into it, eg Crypto, PAV, disk replication?
- Are virtual servers available and can capacity be added dynamically?
- Is there a policy for equipment contention.
So why are customers becoming dissatisfied with the traditional service providers? Cost, or more specifically, value, has certainly been highlighted to us, but it’s not just about the price of a service, and more commonly we are hearing of the current service being inflexible or perceived as such, with time and money being wasted when it takes too long for staff to get through security or be collected from reception, or waiting hours before being given access or logins.
Time can be of the essence during a DR test – It’s quite true that mainframe systems programmers often have a tough time completing DR tests – they normally travel a fair distance, suffer delays obtaining logins and physically getting into the building, then are berated when timescales are constrained and objectives not met.
There are differences and larger players and others could be important to you.
Some providers have an extensive global network of data centres, calling upon expertise in different time zones with the ability for support services to ‘follow the sun’.
Some of the larger providers have a serious level of oversubscription on equipment and seats – just imagine a plane crash in the wrong place inside the M25, who might get to the facilities they need first, and who might be unlucky.
The smaller players will likely provide additional flexibility and more competitive pricing, have processes to address oversubscription risk, and be keen to supply ancillary services such as hardware maintenance to provide a very cost-effective bundle.
For some customers, whether for cultural or political reasons or otherwise, there will always be only a very short list of providers that they will engage with for recovery services. For the majority of users, and for the first time, there are now alternatives offering greater flexibility and value than has been on offer before.
Investigating alternatives can significantly improve your bottom line without sacrificing service levels.
You can find out more about Azamour Solutions, a leading supplier of enterprise class computer hardware, mainframe recovery services and maintenance solutions to corporate and SME clients across the globe, and what they can offer on their Web site at www.azamoursolutions.com.
Find out more about iTech-Ed here.
Source: SANS ISC SecNewsFeed @ February 19, 2017 at 03:09AM