There’s no doubt in my mind that PaaS removes most of the operations tasks associated with running servers. We moved from running on top of IaaS (Amazon virtual machines) to using a PaaS cloud (Tier 3′s Web Fabric, based on Cloud Foundry) and almost all the operations work disappeared.
There is a culture of DevOps growing, and even a culture of so-called ‘NoOps’ where some claim that operations as a discipline can be completely eliminated.
So where does this leave IT operations? Well, it leaves them with a big problem. Because there is still an important role for them to fill, but this role is getting scattered throughout the organization.
Working Around IT
It’s now possible for a business unit in an enterprise to develop and deploy their own solutions. How often does a department face a 12 or 18 month waiting list for IT to build them a solution – or perhaps they can’t make a business case to spend the large amounts it can cost to develop and run software. Where previously they might have built a system out of spreadsheets, emails and people, they can now hire a local software development shop, build the app themselves and deploy it onto a cloud, all with not much more than a credit card.
As is so often the case, IT will seek to provide infrastructure and platforms for apps to be deployed on, and we are indeed seeing a surge in private cloud deployments. But IT really can’t constrain the business any more – they need to make their infrastructure appealing enough that the business will *want* to use it, especially for apps that are not mission critical and where data can effectively live anywhere. If they are restrictive about opening up firewall ports, or slow in providing access to existing systems for new apps, the business can often work around them.
So IT are likely to be faced with a situation where they are being ‘worked around’ and different departments/business units are scattered to the four winds, all using different cloud providers and development shops. And then comes the phone call – “it doesn’t work”. And it’s always IT’s job to fix it, especially when customers are involved.
The Future Role of IT
What IT can continue to provide is being responsible for systems staying up – monitoring, configuration, governance and compliance. Mention ‘governance’ or ‘compliance’ to most developers and they will run a mile (or at least groan). And yet these will continue to be very important areas for companies, and areas where IT has traditionally been the enforcer.
What’s needed is a way to allow the business units to go their own way, choose their own platform providers and suppliers, but then still hand over responsibility for production apps to IT.
We propose that rather than trying to control the activities of developers and business units, IT should embrace the cloud. External platforms and infrastructure, internal private clouds, third party services and internal services can all co-exist. So what’s missing? A way to get a single view across all the applications and services in the enterprise, and a way to monitor and constrain production applications so that IT operations can do their job – making sure this myriad of applications and services stay up, and providing support.
At Appsecute our vision is to provide tools that give developers the freedom to use whatever platforms and tools they want, to have complete control over their development and test environments, and then to still be able to hand over their apps to IT when they go into production.
And then to provide tools that give IT Operations the freedom to control their production environments, free from the interfering hand of developers, and to have complete visibility over the development and test environments as well. In practice this means things like access control, monitoring, escalation, deployment and migration that work across a range of platforms, whatever each developer or business unit happens to be using.
In Tyler’s recent blog post he talked about exactly these features being missing from PaaS – right now PaaS, especially n the public cloud, is a force for chaos in the enterprise. But add a few things and it can massively speed up the development life cycle while still letting IT operations do their vital job.