I think it makes sense to focus on the economy - since we are in unprecedented times. One of the most important technologies as it relates to the economy is data de-duplication. I have been analyzing and writing about data de-duplication since 2003 and in that time it has proven to be real, compelling and valuable to end users. The only reason not to invest in data de-duplication is that you simply don't have that much data. Data de-duplication has been most widely deployed with disk-to-disk (D2D) backup. If you have been considering a D2D backup solution with data de-dupe, then think about actually making it a priority this year. Yes, the stuff really works. If you are just using tape, the economics may not be clear to you because the physical cost per MB of storage for tape is cheaper. But you have to look at the cost of "virtual" MB. The right approach is to measure the cost of the de-dupe solution based on the effective amount of data stored. A 10-to-1 de-dupe solution is backing up 20 TB of data on 2 TB of physical capacity. And achieving 10-to-1 is conservative. And let's face it - tape sucks. You can keep your tape system but reduce the amount of times you use it - instead of daily or even weekly fulls - do your full backups once a month to tape. This reduces tape rotations, lowers operational costs and minimizes human error. Data de-dupe is also finding its way into primary storage as well. IMO this makes a ton of sense. And the good news is that primary and backup de-dupe should be cumulative. But first things first - D2D backup with data de-dupe is a no-brainer. Not that I am implying that if you don't implement it that you have no brain ;-) Okay - for those of you that have objections to data de-dupe for D2D backup or primary - send me some comments and we can discuss these issues one by one. I for one am a big proponent of data de-duplication and believe that it should and eventually will become pervasive. Data de-duplication is really a form of virtualization - virtual data - if you will. Data de-duplication changes the economics of storing data - it reduces power, cooling and floor space consumption - and it enables you to do more than you could otherwise. Data de-duplication is landscape changing and we are still in the early chapters with more to come. But don't worry about how it can help you in the future - look into it now. Data de-duplication solutions should be a high priority for the data center. Go get some.
In the world of IT, there is an interesting balance between innovation and risk. While innovation is what is usually talked about – it is the mitigation of risk that often drives the decision. That is why the IT leaders remain the leaders and the startups spend most of their time pushing boulders up a hill. In the world of IT quantity is quality – the more product you have in the field within production environments – the less risk there is for the next guy. But if that is the case – then aren’t startups always at a disadvantage? Abso-friggen-lutely. Startups have to bring compelling technology to the table in order to even be considered. Startups shouldn’t even be allowed in the door unless that is true. Otherwise it’s a waste of everyone’s time (including their own). But assuming they have done this – then the next most important thing startups need to focus on is risk mitigation! Luckily for the startups – there are end users that are “early” adopters. But I am still shocked at the number of startups that have terrible customer reference programs, awful case studies and very little focus on customer testimonials. And hardly anyone ever, ever, ever leverages their channel partners for testimonials. In many cases, IT professionals are not implementing innovation – that is the unspoken reality – by adopting the leaders. However, they do so because it is the safe choice. And at the end of the day – when it comes to running the operations of your company – you will more often than not go with the proven choice. Taking the “safe” choice isn’t an issue of politics but rather quality – real or perceived. One of the biggest challenges with IT infrastructure is there are infinite variables. Therefore the more a product is implemented – the more common variables it is exposed to – which results in a greater chance of success in your data center. Does that mean that you shouldn’t implement solutions from startups? Yes and no. The focus should actually be on innovation that provides real value to your operations. Here is the litmus test – a startup has to really earn your business by providing something so compelling, unique and valuable that it becomes pretty much a no-brainer. What about the risk? Do your homework and make the startup vendors do their part to mitigate the risk. Even if they have only 10 customers – make sure that you get to talk to as many of those customers as you can. And make sure they give you a case study on each of those customers. Another way you can reduce risk is to use the new product or technology for an application that isn’t mission-critical as the first step. Get a feel for it before you make it a part of your core applications and/or operations.