3 New Year’s resolutions for the cloud in 2018

I’m one of those people who takes time at the new year to define personal objectives for the forthcoming year, some of which I actually achieve. Enterprise IT should be doing the same thing for cloud computing.

Here are my three suggestions for IT’s cloud resolutions for 2018.

2018 cloud resolution No. 1:
Look at your cloud security approach and technology

When I find issues with enterprise cloud deployments in my consulting work, it’s most often around security. Clients often leave aspects of their cloud deployments unprotected or underprotected, and things that should be encrypted are not, while things that should not be encrypted are.  

While I’m not recommending that you gut your cloud security and replace it with what’s cool and new, I am recommending that you take some time to walk through the security solution architecture and ask yourself about where you can improve. Moreover, consider all the security technology in place, what needs to be updated?   What should be replaced?

2018 cloud resolution No. 2:
Look at your cloud training plan

There are two categories of cloud training:

  • Provider training that’s focused on a specific provider such as Amazon Web Services, Microsoft, or Google.
  • General training that provides a good overview of how to make cloud work in enterprises, and all that is involved with that.

You should have a mix of both, as well as some paths for your staff defined to get the skills of a cloud architect, cloud developer, cloud operations specialist, and cloud devops specialist, just to name a few roles. There should be training paths through both vendor and nonvendor  courses to get your staff members the skills they need to perform their duties (which of course must be clearly defined). 

2018 cloud resolution No. 3:
Evaluate your databases

Databases are sticky, and once enterprises have used a specific database, they are not likely to change it. Indeed, what many enterprises have done is just rehost their data on public clouds using the same database they used on premises.

Today we have many options in the cloud, including SQL and non-SQL databases. While there are native databases in public clouds such as AWS’s RedShift and DynamoDB, there are many other options from databases providers that support the public cloud and traditional platforms. Are you using the optimal solution?  

These are just a few suggestions; I suspect that you can name more. Whatever they are, pick a few and follow up. Have a great new year!

Powered by WPeMatico

Cloudops automation is the only answer to cloud complexity

Let’s face facts: We often look at cloud computing as a single place to put our workloads and our data, where magic pixy dust takes care of what’s needed. But today we increasingly understand that the reality of the cloud—like the datacenter before it—is complexity, labor intensitiy, and more costs than we expected.   

This should not surprise anyone. Cloud computing, including the hybrid and multicloud approaches, are complex distributed systems. Although the use of the public cloud buys you not having to purchase and maintain hardware and software, you still have to monitor, manage, and deal with cloud-based systems as if they were down the hall.

In 2018, I think that this is a problem that most enterprises can solve. And we’ll do so using automation, as we have in so many business and manufacturing processes before.

The fact is that most enterprise deal with cloud operations—aka cloudops—using the native tools of their cloud providers. Although that is scalable when you’re just using one public cloud for everything, the reality is that you have to manage traditional systems built within the last 20 years, multiple public clouds, perhaps a private cloud, IoT devices, and data that exists everywhere (with no single source of truth). In other words, a huge mess.  

Automation does not save you from having this mess, but it helps a great deal. 

First, you need to consider the concept. When you automate cloudops, you’re really looking to remove the complexity by placing an abstraction layers between the complex array of systems,and you the person that needs to operate the technology.  

Second, consider the enabling technology. This means using a cloud management platform or other management systems that can automate most of the back-end operations tasks, including backing up systems, putting servers in and out of production, and security operations. The trick is focusing on the broader management technology, and the automation that it providers, versus the cloud-native tools that won’t help you beyond a single public cloud.    

That may seems simple, but it’s not. Today’s cloud management technology does have limits, and there is no magic pixy dust there, either. However, it will get you further down the road by removing the need for people to touc everything all the time, via the use of better and better automation.    

Powered by WPeMatico