Why Server Orchestration?

By Rajat Bhargava Posted April 16, 2014

Music conductor pointing

Sometimes, peers in the industry wonder why we are focused on server orchestration. They also wonder why we don’t just stay focused on user management and security since we built a product that does just that and whether server management is a completely different thing for JumpCloud®.

To answer these great questions, you have to go up to the top level of our vision, mission, and goals for JumpCloud. It’s actually very simple. We aren’t just focused on server orchestration or management; we are focused on reimagining the directory service. And a byproduct of that focus is managing servers.

Very early on, we wanted to make life a lot easier for DevOps and IT folks. Just about every conversation we have with these folks is one where they are completely strapped for time and simply cannot get to everything that they want to do. While we see other parts of organizations busy, we fundamentally believe that the pace of IT dictates the pace of the organization. Therefore, anything that we could do to increase the pace of IT would directly impact the pace of a business. We like that. That’s strategic and a big, meaty problem to solve. Of course, that’s also massive and somewhat diffuse. We looked at this and put it up against what was going on in IT.

Explosion Of The Cloud

The cloud is and has been exploding, and we believe that companies will leverage and adopt cloud. It’s really in its infancy in terms of what can be done with it. To us, that really meant that IaaS was going to be a key component of any IT infrastructure. The key atomic unit for us here was the user. Users need to access a wide variety of different IT resources. Servers are a core part of the engine that any IT organization will need to execute on their business strategy. As a result, we thought about things that we could do to a user’s experience. More specifically and most importantly, we wanted to make DevOps and IT folks move faster with their servers along with their other systems, applications, and networks.

The Server Life Cycle

To understand that, we looked at the server life cycle, which we believe to be provisioning, configuration, monitoring, and management/maintenance. All of these areas need to be automated in order for folks to be able to become faster. The first three – provisioning, configuration, and monitoring – have largely been tackled in the cloud world. On the provisioning side, Amazon, Rackspace, and SoftLayer are doing a great job automating the creation of server infrastructure. Once you have a server spun up, configuration automation tools (seems like there are about a dozen popular ones) can help you deploy everything you need on them. When you add in automated monitoring tools like Nagios and Zabbix, you are in business.

Manual Management And Maintenance Tasks

We also noticed something else. Once things were up and running, there were still lots of manual management and maintenance tasks that needed to be done. Those weren’t automated. Managing users, security, patching, ad hoc reports, and many other relatively straightforward tasks were still largely done manually. We thought it would be useful to start having those tasks automated, so we started with user management and user security. Those tasks are among the first that you encounter with servers, so having a complete user management life cycle would be beneficial.

As we continued working with customers, we kept hearing about more tasks that they wanted automated. They wanted to have a similar concept to GPOs with Active Directory except it should be cross-platform and work in the cloud. So we made a generic task and policy execution engine. Then we would layer on the specific tasks that people wanted: screen lock, disk encryption, disable guest accounts, and others. What if we didn’t create the specific task that somebody wanted? The platform would be open, so they could embed a script or upload an existing script and execute it. It combines the best of both worlds. It has specific functions to solve real problems for our customers but the flexibility to solve more problems through a platform.

JumpCloud Speaks Your Language

That’s the long-winded answer to why we wanted to help IT execute on device management tasks as part of our Directory-as-a-Service® platform. In order to make that happen, we needed to be able to automate the execution of tasks. And that took a server orchestration platform. The beauty of our platform is that you can leverage whatever language you want, including Bash, Perl, Python, Ruby, Go, and Node.js among others. 

Go ahead; automate the way you want. You will not be constrained by learning a specific language or syntax. We let you create tasks or jobs as individual commands or build/upload scripts into our platform. From there, you can execute those tasks against a group of servers. Manually taking a script and executing it on a group of specific servers is time consuming whether it’s done one-off or scripted. You don’t need to take that time anymore, and that’s a huge benefit. Simply apply a tag to your task. It’s executed as you desired on those servers: scheduled or ad hoc.

Part of the reason that things aren’t automated well in the management arena is that you have to take the time to do the next steps. Reporting, auditing, logging, and control all take time if you’re doing it yourself. However, if you have a platform that has those capabilities already built in, then that’s more time saved. The benefit is that you get true visibility into what’s being executed across your infrastructure and by whom. Both pieces of information are critical to any goal of automation.

Secure User Management And Server Orchestration

For JumpCloud, our cloud-hosted directory service combines our centralized user management and security capabilities. It’s a start of a long-term journey to help DevOps and IT folks save time, increase reliability, and provide visibility into their infrastructure. How can we build a service that will save you even more time? What would enable you to move even faster and also allow your company to grow quicker? How can we help put IT out in front and dictating the pace for any company? That’s our goal and our vision. We hope you will join us and help – it’s a big opportunity for all of us. We’d love to hear your feedback.

Rajat Bhargava

Rajat Bhargava is co-founder and CEO of JumpCloud, the first Directory-as-a-Service (DaaS). JumpCloud securely connects and manages employees, their devices and IT applications. An MIT graduate with two decades of experience in industries including cloud, security, networking and IT, Rajat is an eight-time entrepreneur with five exits including two IPOs, three trade sales and three companies still private.

Recent Posts