The DevOps tool business is heating up with competition as the sector itself gets increased attention and traction from a growing number of enterprises.
Providers, such as Puppet Labs, StackStorm, BlazeMeter, CA, HashiCorp and others, are coming to the fore with new agile-use tools, and they quickly are being welcomed into IT shops.
The latest new product in this area comes from HashiCorp, the company behind the Vagrant DevOps development toolset. HashiCorp has released a new product, Atlas, to the general public following a successful beta program.
Atlas, released to general availability on July 8, combines HashiCorp's popular open-source tools for development and infrastructure management to create a version control system for infrastructure. Teams of operators use Atlas to responsibly deploy applications and make changes to infrastructure through a versioned, auditable and collaborative process.
During its beta-testing period, Atlas doubled its nodes under management each month and attracted large enterprises such as Mozilla, Cisco and Capgemini, as well as growing startups, CEO and co-founder Mitchell Hashimoto (at right in photo with The Cube's John Furrier) told eWEEK.
The 6-year-old San Francisco company's goal is to furnish high-quality tools for building a software-managed data center—from development to production, Hashimoto said.
"Hashi means 'bridge,' and we're trying to bridge paradigm gaps and gaps in the future to try and get you in a better place with your data center as a whole," Hashimoto, 25, told eWEEK. "Using our current tools (Vagrant, Consul, Vault and others), we're trying to make these building blocks of a software-defined data center free and open source. We think that's what the industry expects and wants to use, and our DNA is open source, so that's what we do."
Atlas is a commercial edition for building a single management UI, "a single center of automation that takes you from development to production. This is, in fact, powered by our open-source tools," Hashimoto said.
Atlas Combines Several Tools
Vagrant, Packer, Terraform, Serf, Consul and Vault comprise the open-source foundation for Atlas. All the components are hosted and run in Atlas and can be used individually or all together. Here are the components and what they each do:
--Vagrant, for managing development environments;
--Packer, which automates building artifacts, such as Amazon Machine Images (AMIs), Docker containers, OpenStack images, VMware images and more. Packer-produced artifacts can be stored in Atlas.
--Terraform automates infrastructure provisioning on Amazon Web Services, Google Compute Engine, OpenStack, Azure and more.
--Serf's gossip protocol is used by Consul for failure detection and message propagation in a server cluster.
--Consul is a tool for runtime orchestration, complete with service discovery, a key/value store and health checks.
--Vault manages keys and secrets in distributed systems and is used to store all credentials used by Atlas.
When all the components are used from development in Vagrant to production monitoring with Consul, Atlas is a fully versioned application delivery pipeline. Teams of operators can review and collaborate on Vagrant development environment versions, server configuration updates made by Packer, infrastructure updates made by Terraform and alerts triggered by Consul.
This simplifies rollbacks, makes infrastructure debugging feasible and allows teams to easily share configurations, Hashimoto said.
DevOps Among Hottest Trends in IT
DevOps (a mashup of the terms "development" and "operations") has come to the fore as a relatively new business process aimed at overcoming complexities that IT, systems administrators and developers face on a daily basis. Even though some organizations are implementing DevOps practices, many still don't completely understand what DevOps is; they either are resistant to or don't realize the benefits of deploying this new approach.
DevOps is a set of methods, principles and practices for collaboration and integration between development and IT operations. This collaboration increases agility and reduces friction between development and operations, resulting in faster software/application deployment and quicker problem detection.
The goal of operations and infrastructure management is to deploy and maintain applications with an automated, error-free and auditable process. However, many organizations use manual processes that are time-consuming, error-prone and difficult to scale. As more technologies are added to the modern data center, from configuration management to containers and schedulers across multiple clouds, the management process becomes more complex and risky.
Mozilla, for one prominent example, is using Atlas as the infrastructure management layer for its Socorro project, which recently migrated from a private data center to the cloud. Mozilla is a user of the open-source projects Packer, Terraform and Consul and is now tying them together through Atlas.
Cisco is using Atlas components to power Shipped, its new developer platform (currently in alpha) to bring applications from development to production in five minutes.
Many Different Use Cases
Capgemini built Apollo, its next-generation application delivery platform on Atlas and Mesos. The previous platform relied on manual processes for server configuration and provisioning, which led to a fragile, time-consuming deployment pipeline and information silos between teams.
A hosted version of Atlas is available now, and on-premise installations will be deployed in late 2015. Go here to sign up for Atlas or request on-premises information.
Pricing is $40/node/month with the first 10 nodes free. Development features, which are free, include Vagrant box builds and hosting and Packer builds and artifact storage, Hashimoto said.