In a recent discussion with a colleague, he looks after one of our Strategic Enterprise Accounts, we got chatting about some of the ways that his account is looking to reduce costs. The one that most interested me was in Test and Development. Their view was that due to the nature of this environment being less critical than production they could get away with much lower cost / commodity style storage and save a bit of money.
This really stuck with me as it seems to be symptomatic of a major issue within IT in many companies today, often large ones. Bold visions for IT are often crafted and communicated at the highest levels but as this filters down to the various groups and functions within IT it seems to get more and more diluted, to the point where people default to traditional behaviour. It’s just too easy to measure cost reduction and get that warm feeling that if you’ve saved a bit of money you’ve somehow been successful in your job.
Personally I think this is a short sighted view. This company is trying to move toward an Agile Project development methodology, which is predicated on the ability to rapidly clone and test, clone and test. Their existing NetApp storage has the ability to do this. FlexClone is a phenomenal feature which allows for the near instant creation of zero overhead copies for exactly this purpose. That’s right, instantly available and no additional storage needed, and what’s more this capability could be given to the people that need it need another test and development environment? simply click and you’re done!
Now what if you could extend this out to the likes of Amazon. If someone needs a new environment for Test and Dev then why buy the infrastructure at all? Why not use our Cloud Manager software to enable an AWS instance to be created inside Amazon with Cloud ONTAP providing the underlying virtual storage layer? That way your AWS test environment looks, feels and behaves the way your production ONTAP storage environment does. Its also now incredibly simple to get data in and out using our SnapMirror software to just replicate the data across.
It’s worth considering that for these kind of rapid Test and Dev activities the cheapest storage is probably the storage you don’t own yourself.
There are two measures here. If you’re creating clones on our storage or using AWS instances, instead of full copies on more commodity storage then somewhere there will be a crossover point where the inefficiencies of commodity mean that you 
actually end up paying more. But I don’t want to focus on this, what I want to focus on is the impact that these technologies could have on the entire Agile process. How much value would be created for the teams that currently sit there waiting for environments to be setup if it could be done instantly, time and time again? This is one of the reasons we’re seeing the continued growth in ‘Shadow IT’ whilst this IT group is focused on trying to reduce costs and not working with the Test and Dev teams to find better and faster ways to do things, the Test and Dev teams go and find alternatives and run into the arms of Amazon or Microsoft or any one of many many Cloud Providers that have designed with these types of consumption models in mind. Recent estimates now state that 30% or more of the money that companies spend on IT is spent outside of IT and this number is rising fast..
Want to know what’s possible? Then start by looking at our recent Epic Story with Symantec, who by using these technologies saved 700 years of engineering effort, yes 700 YEARS!
I guess the question to ask yourself really is…
Do I want to save a little bit of money? or do I want to make a lot?
Matt,
I could not agree with your assessment more. I have visited many customers in the last 45 days where the sentiment was to move to an Agile development process, the issue was, the processes they had behind trying to get there were 1) costing them more money, 2) putting more of a burden on IT resources and 3) not allowing development to meet the SLAs they wanted in moving to an Agile process in the first place.
I also agree that when customer put forth RFP’s for storage solutions, they make sure that the solution they choose has all the storage services known to man; compression, deduplication, thin-provisioning, tiering, and yes snapshot capabilities. Then, they purchase the array, and only use 50% of the features available to them, that if they leveraged the technology, could actually save them money in the long run.
We see time and time again that in order for IT to service the SLAs of any LOB (development, finance, marketing, etc…) they tend to default to making copies of their data for these LOBs by doing full recoveries from a backup. This process is time consuming, eats a lot of system and network resources, causes data sprawl (as nothing ever gets cleaned up) and does not meet the businesses SLAs (especially for an Agile development model.
At Catalogic Software we have put a lot of thought into this use case in particular – Test/Dev –> DevOps. The ability to leverage the NetApp Snaps, Vaults, Mirrors, as FlexClones, in ANY location, including the cloud (thanks to NetApp’s Cloud ONTAP) means a business can effectively have an Agile development environment(s) in minutes (not weeks, months), reduce the burden on IT and meet the business requirements. Catalogic’s ECX software in fact helps IT visually get a handle on the creation of their Snaps, Vaults, Mirrors, understand where they are and where they are needed, automates the creation of these services and then orchestrates bringing these data sets up, in many locations, to be used for multiple business operations such as DevOps. In addition, ECX can be scheduled to tear down last nights environment and ensure a new ‘clean’ environment is set up the next day.
The ECX technology has enabled many customers to move from traditional, costly, Test/Dev to a robust DevOps, Agile environment, leveraging NetApps technology, technology that you have already paid for.
Nice piece Matt, good luck.
Steve