Where is your information? Would you describe it as being in your infrastructure or your data infrastructure? Let’s look at what data infrastructure is, why it is important, and the specific characteristic of reliability availability serviceability (RAS). Then we will close by reviewing some common issues and advice.
When you use the Internet, whether for business or personal reasons, you are fundamentally reliant upon data infrastructures. A data infrastructure is a backend computing concept (it is backend computing, essentially), so it is understandable that it is often called by names that don’t fit it quite as well – such as simply infrastructure, or the data center.
Infrastructure is a term that is used for the set of tools or systems that support various professional or personal activities. An obvious example at the public level, in terms of infrastructure that is maintained by the government, is the roads and bridges. These elements are the basic structures through which people can store, contain, or transfer themselves, products, or anything else – allowing them to get things where they otherwise couldn’t.
Infrastructure is, in a sense, support to allow for the possibilities of various functions considered essential to modern society. In the case of IT services specifically, you could think of all technological components that underlie that tool as infrastructure, noted Greg Schulz in Network World.
The basic issue is that infrastructure is an umbrella term for these functional, supportive building blocks. There are numerous forms of infrastructure that must be incorporated within an information technology (IT) ecosystem, which are best understood as layers. The top layer is business infrastructure (the key environments used to run your business). Beneath that layer is the information infrastructure, the software, and platforms that allow the business systems to be maintained and developed. Finally, beneath the information infrastructure is the data infrastructure, as well as the actual data centers or technological habitats. These physical environments can also be supported by outside infrastructure, especially networking channels and electricity.
Understanding the context in which data infrastructure (ranging from cloud hosting to traditional onsite facilities) exists, let’s explore the idea in its own right.
If you think of a transportation infrastructure as the components that support transportation (the roads and bridges) and a business infrastructure as the tools and pieces that support business interactions directly, you can think of data infrastructure as the equipment or parts that are there to support data: safeguarding it, protecting it from destruction, processing it, storing it, transferring it, and sending it – along with the programs for the provision of computing services. Specific aspects of data infrastructure are physical server machines, programs, managed services, cloud services, storage, networking, staff, and policies; it also extends from cloud to containers, from legacy physical systems to software-defined virtual models.
Purpose: to protect and to serve (the data)
The whole purpose of your data infrastructure is to be there for your data as described above – protecting it and converting it into information. Protection of the data is a complex task that includes such concerns as archiving; backup and restore; business continuity and business resiliency (BC/BR); disaster recovery (DR); encryption and privacy; physical and logical security; and reliability availability serviceability (RAS).
It will often draw some amount of attention when a widely used data infrastructure or application environment goes down. Recent outages include the Australian Tax Office, Gitlab, and Amazon Web Services.
The troubling thing about the downtime incidents that have been seen in these high-profile scenarios, as well as ones that were not covered as much outside of security circles, is that they are completely avoidable with the right safeguards at the level of the software and the data.
A large volume of disasters and other unplanned downtime could be reduced or eliminated altogether, said Schulz in a separate Network World piece. “[I]f you know something can fail,” he said, “you should be able to take steps to prevent, isolate and contain problems.”
Be aware that there is always a possibility of error, and any technological solution can experience a fault. People worry about the machines. Oh, it is easy to point fingers! However, the greatest areas of vulnerability are the situations in which humans are determining and controlling setup of computers, applications, plans, and procedures.
What is the worst-case scenario? If data loss occurs, it can be complete or partial. Features of data protection are both vast and granular, having to incorporate concerns related to diversely distributed locations, data center facilities, platforms, clusters, cabinets, shelves, and single pieces of hardware or software.
What is RAS?
Why must your data infrastructure have RAS? Reliability, availability, and serviceability is a trio of concerns that are used to architect, build, produce, buy, or implement an IT part. This concept was originally “deployed” by IBM when the company wanted to come up with standards for their mainframes; at that point, it was only a characteristic pertinent to hardware. Now RAS is also used to describe applications, networks, and other systems as well.
- Reliability – the capacity of a component of hardware or software to meet the specifications its manufacturer or provider describes.
- Availability – the amount of time that a computing part or service works compared to the entire time that the user expects it to work, expressed as a ratio.
- Serviceability – the extent to which a piece of a computing environment is accessible and modifiable so that fixes and maintenance can occur.
Top issue #1: letting software take the lead
Both data infrastructures and threat landscapes are increasingly software-defined. Today is the era of the rise of the software-defined data infrastructure (SDDI) and software-defined data center (SDDC). In this climate, it is good to start addressing the field of software-defined data protection – which can be used to allow for better availability of your data infrastructure, along with its data and programs.
In today’s climate, the data infrastructure, its software, and its data are all at risk from software-defined as well as traditional threats. The “classic” legacy problems that might arise are still a massive risk; they include natural disaster, human error, glitches in programs, and problems with application setups. Software-defined issues run the spectrum from spyware, ransomware, and phishing to distributed denial of service (DDoS) and viruses.
Top issue #2 – getting ahead of the curve
We should be hesitant to shrug it off when there is a major outage. We should ask hard questions, especially the big one: “Did the provider lower operational costs at the expense of resiliency?”
4 tips for improved data infrastructure
Here is how to make your data infrastructure stronger, in 4 snippets of advice:
- You will want to consider the issue of resiliency not just in terms of cost but in terms of benefits – evaluating each of your systems in this manner. The core benefit is insurance against outage.
- Create duplicate copies of all data, metadata, keys, certificates, applications, and other elements (whether the main system is run third-party or in-house). You also want backup DNS so you cannot effectively be booted from the internet.
- Data loss prevention starts at home, with a decision to invest in RAS or to lower your costs. You should also vet your providers to make sure that they will deliver. Rather than thinking of data protection as a business expense, reframe it as an asset – and present it that way to leadership.
- Make sure that data that is protected can also be quickly and accurately restored.
Simply by thinking of your programs and data on a case-by-case basis, as well as implementing strategies (such as deduping, compression, and optimized backups) to minimize your data footprint, you will spend less money while creating better redundancy. To meet this need, your data infrastructure should be as resilient as possible – but also be attached to incredible support so you can quickly adapt.
Are you rethinking your data infrastructure? At Total Server Solutions, we believe strongly in our support – but, as they say, don’t just take our word for it. From drcreations in Web Hosting Talk: “Tickets are generally responded to in 5-10 mins (normally closer to 5 mins) around the clock any day. It’s true 24/7/365 support.”
We’re different. Here’s why.