Adaptive Self-Service Cloud Option Helps RingCentral Optimize Big Data and Complex Systems

IDC 1875
I D C C U S T O M E R S P O T L I G H T
Adaptive Self-Service Cloud Option Helps
RingCentral Optimize Big Data and
Complex Systems
March 2015
Sponsored by Qubell
Introduction
RingCentral Inc. is a publicly traded provider of
software-as-a-service (SaaS) communications solutions
for businesses. Delivered via a state-of-the-art cloud
infrastructure, its phone and communications systems can
offer customers a flexible and cost-effective way to support
distributed workforces, mobile employees, and the
proliferation of "bring your own" communications devices.
Its flagship service, RingCentral Office, is an enterprise
solution that allows customers and their employees to
communicate via voice, text, and fax on multiple devices,
including smartphones, tablets, PCs, and desk phones.
Founded in 1999, RingCentral is headquartered in San
Mateo, California, and has offices in Denver, Charlotte,
London, and Xiamen, China.
RingCentral has experienced rapid growth in the past
several years as businesses redefine how they work and
increasingly seek integrated cloud communications solutions
that can be used across the enterprise. With a large and
growing customer base, the cloud-based provider deals daily
with the rapid deployment of complex, heterogeneous
software systems and the related challenge of configuration
management. Although the company built internal tools to
address this challenge, many hands still needed to touch a
component before it went to production, resulting in a
manual process that was both inefficient and error prone.
RingCentral wanted to increase efficiency with automation
and self-service for faster testing and application releases
via dynamic and rapid provisioning and also to decrease
errors by minimizing human intervention in software
deployment and configuration management. The company
began looking for an automated solution suited to its
complicated software environments and found it in
application deployment and configuration management
tools from Qubell Inc.
Solution Snapshot
Organization: RingCentral is a publicly
traded provider of cloud-based phone
and communications systems for
businesses.
Operational Challenge: RingCentral
needed to solve configuration
management problems for rapid
delivery of changes and quality for
complex Hadoop big data analytics and
other systems on which the company's
phone and communications business
depends.
Solution: The company selected
Qubell Inc. and its autonomic
application management platform.
Project Duration: Initial big data
deployment took 6 months (rather than
12+ months). Other deployments are
ongoing and will be completed in 1Q15.
Benefits: The company moved from an
outdated version of Hadoop and gained
application speed, quality, and reliability
with Qubell self-service. Teams spin up
tests in days, not weeks, and deploy on
Amazon cloud for agility in a dynamic
market. Data analytics enable insight
into phone and communications
performance and usage.

©2015 IDC2
Implementation
RingCentral has customers in the United States, Canada, and the United Kingdom in industries such
as healthcare, financial services, legal, retail, and real estate. The company has successfully
developed a brand cache, further differentiating its offerings and increasing brand awareness in an
already crowded market serving small and medium-sized businesses (SMBs). However, companies
live or die today based on how well they innovate and deliver services to their customers — and
RingCentral is no exception.
To engage its customer base with innovative communication services, RingCentral requires both
speed and agility to deliver new releases quickly, frequently, and reliably. Given the nature of its
business, the company is in a state of constant deployment with frequent major new releases. "We're
in a really agile environment, and so this problem never goes away," says Kira Makagon,
RingCentral's Executive Vice President of Innovation, who is responsible for product strategy, product
management, engineering, and operations. "It only gets worse if you don't manage it."
RingCentral's homegrown solution for configuration management and software deployment was no
longer robust enough to support the company's increasingly complex business and production
environment. "There are many variables that go into our configuration," says Makagon. "We have lots
of components — over 200 that get deployed — and they are in different categories. Assembling all
of that in a standardized fashion while minimizing the amount of human intervention is the challenge."
Although the company ran faster than competitors, its processes were inefficient. And, says
Makagon, "Inefficiency means errors and bottlenecks." The company needed an automated solution
that would help it implement and manage changes to its complex online services rapidly and without
risking stability.
RingCentral put together a team from engineering and operations to evaluate products that address
configuration management, quality, and deployment orchestration. However, the team found few
tools with the flexibility and functionality required. "We looked at a number of products, but they
weren't flexible enough for us," says Makagon. "They just could not adapt to our environment. We'd
have to redo quite a bit of what we do, and that's very expensive."
Enter Qubell Inc., a provider of application-centric environment and configuration management
solutions. The company's application management platform can enable more adaptive applications
with self-managed services to help configure and optimize the applications in response to changing
and dynamic environments. Customers can use Qubell to create a cloud platform for integration
testing and production deployments that can reconfigure themselves as changes come through the
application pipeline from development to production.
An advantage of Qubell's solution is its ability to fit into and work with an existing environment. Thus,
RingCentral was able to leverage existing assets and avoid reworking what it had already built.
"Qubell's tool lets you plug in your own scripts and parameter configuration. It is flexible about where
it points to so you can deploy where you need it. Their model also allows lots of substitution and is
easily changeable," notes Makagon. "It's modular enough where it's easy for a company like us to
manage exceptions than it would have been with other tools."
To evaluate the platform's effectiveness, RingCentral brought Qubell in for proof-of-concept pilots.
One major project completed last year enabled a big data deployment; other projects will include
other infrastructure components.
The project manages RingCentral's big data deployment, which is both internal and customer facing.
Leveraging Qubell, this initiative automates a process that used to be done manually: updating cost
centers and managing and rolling out new scripts to integrate and interface with various elements of
the onboarding system. Qubell enables self-administration for both development and data science
teams to automate this deployment process.

©2015 IDC 3
RingCentral's Senior Data Engineer Michael Becker was brought in to set up the company's Hadoop
big data environment; the company was using an outdated Hadoop version at the time, and support
was expiring. To enable data analytics and visibility into customer phone and communications
performance and usage in a way that was cost effective, the company needed to move quickly to the
most current release of Hadoop. Leveraging Qubell's self-service, autonomic capabilities, RingCentral
was able to set up cloud-based testing and orchestration environments for its complex big data
system in a matter of six months. Doing so without Qubell would have taken at least twice as long,
Becker says, and could have been challenging to accomplish at all. The system is used across four
countries by around 20 IT team members.
"We needed the ability to have a sandbox where we could do development, and a sandbox for Hadoop
is not a trivial matter because you have so many nodes that are involved. We needed a turnkey
system," says Becker. "Now we are able to dynamically create and destroy those Hadoop clusters on
Amazon AWS for our sandboxes. And that's also good because we have the flexibility of self-service.
Over the 2014 holidays, we were able to destroy those clusters and save ourselves money on AWS."
As a result of the project, RingCentral is loading onto a single platform the call records landing on its
Hadoop infrastructure and additional call detail records from other systems. The company is now able
to cross-correlate the data for deeper insight into how customers are using their RingCentral phone
systems. This also allows customers to understand usage patterns, what worked, and how they are
leveraging their services to help assess the benefits that they're accruing from RingCentral.
Challenges
RingCentral is a global enterprise whose workforce is widely dispersed geographically. The
production team and developers are distributed worldwide. When it comes to an implementation,
RingCentral not only has to decide how and where to roll out a solution but also has to deal with the
challenge of communicating this information to its distributed teams.
However, Makagon says that the hardest thing to manage has been the expectations unleashed by
implementing new tools. "Developers are pretty finicky, and they all have their things that they want.
Now that they have a chance to switch tools, there's a whole bunch of 'Can we get this done?' and
'Can we get that done?' They all have ideas about it."
Benefits
The Qubell system is an integrated orchestration environment. This meant that RingCentral's teams
didn't have to take different components and put them together themselves for the big data project. The
self-service component is a major benefit, according to Becker. "If I assign a task to a developer to
modify some code on Hadoop, they could self-service a Hadoop cluster, and it will deploy all our settings
and also deploy the code to the existing trunk or branch on that Dev cluster. Then the developer will be
able to do their development, post the code back to the branch, and destroy the cluster."
Other teams without Qubell have to wait for operations to deploy the hardware and operating system
and install the software, which can take two to three weeks compared with one day. Qubell enabled
RingCentral users to set up their sandboxes for Hadoop in days rather than weeks.
"We're able to be very agile. So, if we want to upgrade some components onto our Hadoop cluster,
we can spin up two versions of a cluster and have two different groups working on two different
branches, and we don't have resource contention between them ... and I'm able to determine the
optimal infrastructure for what it is we're trying to do," says Becker. "That gets people excited.
We don't have to get into the details of building out infrastructure, and the ability to spin it up and
down allows us to not even think about that." He says this lets his teams focus on what's important
and differentiating.

©2015 IDC4
Qubell's consistency across environments creates efficiencies of scale because it automates scripting
and minimizes human error, he adds. Configuration management is also a key benefit. "All our
configurations are stored in Qubell, which is a core part of the full software development life cycle and
enables application deployment," says Becker. Because this big data initiative was a relatively new
project, it was easier to automate.
A harder, separate Qubell project at RingCentral involved deployment of the company's media servers.
This project was challenging because it touched more people and involved software that had been
around longer with associated processes and history, but it was able to be accomplished with Qubell.
Other deployments involving Qubell will be completed during 1Q15, so it is too soon to quantify actual
ROI. However, Makagon sees primary paybacks in Qubell's ability to enable speedier deployment
and to help reduce error rates. "Humans are much more likely to make mistakes," she says. "And
these mistakes can be really costly in the production environment." Makagon expects another benefit
of automating application deployment and configuration management to be efficiency and a potential
staff reduction of around 10% for those manual tasks, with a related retargeting of teams to
differentiating areas for the business and/or staff cost savings.
Benefits resulting from the big data project were very significant because the team previously had to
administer scripts, manage deployment, and then manage rollbacks manually if anything went wrong.
This involved huge team overhead (writing code, algorithms, and managing the data). The
homegrown tools that the team was using before were not as robust and required constant work,
which slowed implementations (a major challenge). And other commercially available tools that the
team evaluated didn't offer the flexibility and robustness of Qubell, according to Makagon. She
recommends that companies making the transition to automation do a full proof of concept because
tools can seem great initially, but when taken to the next level, they can be rigid and not offer depth of
needed functionality.
Qubell's service also makes possible new capabilities for the business. "It will help us with global
deployments for better life-cycle coordination with configuration management, quality, and faster,
continuous release cycles," says Makagon.
Methodology
The project and company information contained in this document was obtained from multiple sources,
including information supplied by Qubell, questions posed by IDC directly to RingCentral employees,
RingCentral's corporate Web site, IDC research, and Bloomberg Business.
A B O U T T H I S P U B L I C A T I O N
This publication was produced by IDC Custom Solutions. The opinion, analysis, and research results presented herein
are drawn from more detailed research and analysis independently conducted and published by IDC, unless specific vendor
sponsorship is noted. IDC Custom Solutions makes IDC content available in a wide range of formats for distribution by
various companies. A license to distribute IDC content does not imply endorsement of or opinion about the licensee.
C O P Y R I G H T A N D R E S T R I C T I O N S
Any IDC information or reference to IDC that is to be used in advertising, press releases, or promotional materials requires
prior written approval from IDC. For permission requests, contact the IDC Custom Solutions information line at 508-988-7610
or gms@idc.com. Translation and/or localization of this document require an additional license from IDC.
For more information on IDC, visit www.idc.com. For more information on IDC Custom Solutions, visit
http://www.idc.com/prodserv/custom_solutions/index.jsp.
Global Headquarters: 5 Speen Street Framingham, MA 01701 USA P.508.872.8200 F.508.935.4015 www.idc.com

Adaptive Self-Service Cloud Option Helps RingCentral Optimize Big Data and Complex Systems

More Related Content

What's hot (18)

Viewers also liked (20)

Similar to Adaptive Self-Service Cloud Option Helps RingCentral Optimize Big Data and Complex Systems (20)

Recently uploaded (20)

Adaptive Self-Service Cloud Option Helps RingCentral Optimize Big Data and Complex Systems