Latency and jitter fall 70 percent with ContainerFlow; response times improve 51 percent—benchmarking study.
SEATTLE — Dockercon — June 21, 2016 — Developers writing applications with microservices tools like Docker containers and the Kubernetes container cluster management system today gain a new tool built for the DevOps era, resulting from a collaboration announced in March between AppFormix and Intel(r).
ContainerFlow is a cloud service optimization tool that provides advanced monitoring, scheduling and performance management for DevOps environments, where containers and VMs can have life cycles much shorter than in traditional development environments. ContainerFlow is currently the only tool that integrates Intel Resource Director Technology to provide major improvements in application performance, especially with container-based applications.
The collaboration that led to today’s product announcement kicked off via a joint announcement between the two companies at Intel Cloud Day in March.
In a solution brief, “Realize The Performance Of Your Cloud Infrastructure,” testing shows improved response times averaging 51 percent with ContainerFlow. The testing also shows an average 27 percent improvement in throughput. By actively managing cache allocation in real time, ContainerFlow reduces latency and jitter by 70 percent.
Importantly, the variances of the average and peak latency are significantly reduced. Peak latency is reduced by 70 percent to 146 milliseconds, and the standard deviation of peak latency is 17 milliseconds, compared to 95 milliseconds without ContainerFlow.
ContainerFlow works out of the box for microservices built using Docker containers, using Kubernetes an open source system for managing containerized applications.
*** More at: http://www.appformix.com/intel/ ***
ContainerFlow is the first scheduling and monitoring tool built specifically to address the unique needs of cloud-native applications and DevOps environments. Applications built with containers and microservices run in environments where many thousands of containers are created, used and discarded in only a few minutes. ContainerFlow works in real time, delivering a critical improvement over monitoring tools built for traditional, monolithic applications.
Before ContainerFlow, microservice application developers and infrastructure operators were stuck with old tools designed for those monolithic technologies, where it is common for schedulers to take several minutes to do their work. With containers and microservices, however, the actual lifetime of processes being scheduled is less than the amount of time legacy tools need to schedule and monitor performance.
“The next generation of cloud depends on the industry’s ability of create technologies that make apps built on microservices easy to write, deploy, scale and iterate,” said Sumeet Singh, CEO and founder of AppFormix. “Conventional tools take longer to do their jobs than the lifespan of the services they’re scheduling and monitoring. Effective tools cannot be more complex and time-consuming than the problems they are trying to solve. Current tools were begging for replacement, so we built ContainerFlow to give users the experience they expect.”
Customer and Partner Quotes
“When isolation is imperfect, containers cannot prevent interference in resources the operating-system kernel doesn’t manage, such as level 3 processor caches and memory bandwidth,” said David Aronchick, Senior Product Manager for Kubernetes. “The approach taken by AppFormix offers hope for operators struggling with managing performance in container-based environments reliably, by giving them a tool designed for that use case.”
<INTC QUOTE IN APPROVAL>
“With AppFormix, Rackspace gets a team that’s solely focused on solving scheduling and management issues that arise in a world where the lifecycle of compute instances is very short,” said Bryan Thompson, senior director, product management of OpenStack Private Cloud at Rackspace. “The Appformix software aligns with our mission to help customers remove the complexity of deploying and operating OpenStack Private Clouds while driving benefits for their business.”
“Our applications demand reliable performance,” said Joshua Barry, product director, cloud at ViaSat. “The combination of technologies in Intel RDT and AppFormix ContainerFlow has the potential for our operators and developers to better understand and manage cloud environments, delivering better user experiences and infrastructure efficiency.”
- Watch the video about AppFormix and Intel(r) RDT
- Read why maximizing CPU availability isn’t enough <NEED LINK ONCE BLOG LIVE>
- Grand Award winner at Interop
- Rackspace license deal with AppFormix
- Borg, Omega, and Kubernetes blog post by the Kubernetes team at Google
AppFormix is a startup delivering a cloud service optimization platform targeting enterprise devops teams. Via real-time analytics, capacity planning, and orchestration tooling, AppFormix software gives cloud operators and applications developers a shared dashboard to understand performance across the entire cloud stack, for hybrid, private, and public enterprise clouds. AppFormix works with popular cloud environments like AWS, OpenStack, Azure and Kubernetes. AppFormix software captured the Grand Award winner at Interop 2016 and was named a technology to watch by the MIT CIO Symposium. The company is backed by August Capital and is headquartered in San Jose, Calif. More atwww.appformix.com.
Cathey Communications for AppFormix