A couple of years ago tech executives at FICO wanted to update their infrastructure. “OpenStack seems to be the wave of the future, so we gave it a run,” says Donald Talton, senior manager of platform operations and cloud engineering at the credit rating agency.
SolidFire Donald Talton
FICO considered using VMware, but felt that the “momentum” of OpenStack was stronger, Talton says. And so began FICO’s use of OpenStack’s IaaS open source private cloud software. Talton says it’s been great, though that doesn’t mean it’s been easy.
+MORE AT NETWORK WORLD: OpenStack fends of Amazon, Microsoft and VMware | Cool products at OpenStack Austin Summit +
OpenStack officials will be the first to admit that the open source cloud project is still maturing. “We’ve done a good job delivering a robust set of infrastructure services, now we need to make sure it’s a really great experience, not only for the cloud operators, but for the end users, too,” says OpenStack Foundation Executive Director Jonathan Bryce. A survey of OpenStack users found that complexity of deploying and managing OpenStack clouds continues to be a pain point.
For Nate Baechtold, IT architect at EBSCO Publishing, which builds search software for libraries and universities, deploying OpenStack was not a major challenge; adjusting to managing it was. Vendors such as Red Hat have made the process of spinning up OpenStack clouds on Day 1 as easy as possible. “Day 2 is where the real work begins,” Baechtold says.
EBSCO uses an OpenStack distribution from Red Hat for its private cloud and Avi Networks for software-based load balancing. EBSCO also runs some workloads in Amazon’s public cloud. Baechtold said one advantage of building an OpenStack private cloud is that it operates similarly to AWS. Users within EBSCO can request and spin up virtual machines, and there is centralized management of them; it’s a very similar model to using the public cloud. Even though EBSCO uses both clouds, Baechtold says integrating OpenStack with AWS is not high on his priority list. The clouds are used for different purposes, most of which don’t overlap.
FICO also uses Red Hat’s distribution of the OpenStack code, combined with Ceph for standard storage and SolidFire for low-latency (2 to 3 millisecond response time) storage. That’s helpful for when a customer requests a credit check on a consumer and needs an answer in 15 seconds, Talton says.
FICO uses many of the core open source components that make up OpenStack: Cinder block storage code to orchestrate storage; Neutron to manage the Cisco networking gear; and Heat for automatic deployment.
Baechtold and Talton agree that a key to successfully deploying OpenStack has been building up in-house expertise of the open source code. Even with the support of vendors, Baechtold says, “We need people skilled enough to run this thing.”
Talton says one of the biggest challenges is keeping up with the rapid pace of change in the open source community. A new release of OpenStack is published twice a year. The core projects – compute, network and storage – are fairly stable and simple; but new emerging projects can be a challenge to integrate.
How does he deal with it? “I’ve got good engineers,” he says. “If you’re going to run production OpenStack at real scale, you need full stack engineers.”
OpenStack this week launched a new certified OpenStack Administrator exam to help build up skills of cloud architects. But the challenge of finding OpenStack talent and training cloud enthusiasts on the platform remains.
