It’s OSCON week so it’s no surprise that a number of interesting open source-related announcements are happening this week. Yesterday it was Canonical announcing that it was crowdfunding the development of a mobile phone (yes, many of us are scratching out heads over that one) and today it’s the turn of CloudMunch who are announcing a DevOps solution for OpenStack. First a little intro to CloudMunch - the focus of the company is to democratize DevOps by providing a DevOps Management Platform. The idea being that no matter what particular tools a development team uses, they can manage them all from one place. Specific functionality that CloudMunch brings includes:
- Project Level Self-Service Provisioning lets project teams control the usage and management of infrastructure at the project level rather than having operations do this at organizational level.
- DevOps App Builder makes possible rapid creation and deployment of new apps, integrations, tests or deployments. Apps are discoverable, context specific, and ready for self-service usage.
- Customizable Software Delivery Progression Workflow ensures that application and infrastructure software builds go through all checks and balances in a transparent way from development, test to release.
- Integrated Open Dashboard gives code progression project visibility and alerts to project managers, testers, operations, and developers.
- Cloud Scale DevOps Engine enables deployment and continuous integration with enterprise class scalability and performance, enabling master-master and master-slave clusters.
CloudMunch already provides a plug-n-play framework that integrates with other open source tools such as Jenkins, Chef, Selenium and others. It makes sense to include deep infrastructure support on a platform such as this as it enables a consistent and end-to-end approach to the development/operations overlap. Hence CloudMunch’s timely announcement of OpenStack support. The gist of the news is that the integration allows developers to use continuous integration and delivery when deploying applications on top of OpenStack – as such it allows the flexibility of an IaaS offering, with some of the visibility of a PaaS solution.
It’s an interesting fact that PaaS uptake has been less rapid than many of us would have liked. While it’s hard to deny that PaaS offers significant value in terms of allowing development teams to focus on core competency and forget about infrastructure, there has still be some reluctance to adopt services further up the stack.
Given this fact, a solution such as CloudMunch offers makes sense – if a team can’t remove the need to handle infrastructure-level operations, then at least they can have one central place to gain visibility over the different tools. The purist in me says it’s sub optimal, but the pragmatist in me says it’s a logical reflection of enterprise’s lack of comfort with PaaS.
The big question here is how long it will take until enterprises gain confidence that PaaS is here to stay – my beilief is that it will happen sooner rather than later in which case tools such as Cloud munch become less important – they’re still undoubtedly have market opportunities – but they’ll be slimmer than under an IaaS-only paradigm.
Time will tell, but in the meantime CloudMunch for OpenStack looks like a great interim step.