Video Screencast Help
Symantec Appoints Michael A. Brown CEO. Learn more.
Cloud Platform Engineering

Building Symantec’s Cloud Platform on OpenStack

Created: 10 Jun 2014 • Updated: 21 Jun 2014
knewstadt's picture
+2 2 Votes
Login to vote

I know what you’re thinking…  why am I talking to you about Symantec... cloud...  and OpenStack?

When people think of Symantec, their minds often go straight to antivirus.  In fact we do a lot more than that, we secure and manage information.  Drill in a bit, and you’ll find that this encompasses lot of things: big data analytics, massive data storage, real-time traffic scanning, management of 100s of millions of endpoints….  It’s hard to list everything we do, but the point is that what we do requires a huge amount of infrastructure.  And we’re currently in the process of building a new cloud platform, where these and all other Symantec products and services will live.

We’re building this new platform on OpenStack.  As we go through this process we occasionally stumble over issues and obstacles, sometimes frustrating but usually interesting and fraught with possibility.  You might be discovering the same ones, or maybe you’ve missed the ones we’ve found and discovered others instead.  The fact is that we’re all working through this together, and we want to share our thoughts and experiences with you.

This posting will be the first of many.  We’d love to hear your feedback, and if you’re interested in participating in the discussion please reach out.  Here are some questions we’re asking ourselves, that maybe you are also asking yourself, and that we’ll be talking about in this blog…

 

How Do You Deploy and Operate OpenStack at Large Scale?

You’re running thousands of physical nodes, many data centers running active-active, complex multi-tenancy.  How do you build it out and operate it?  How do you know you’re getting the most out of your infrastructure, tuned for the best performance?  For example, we presented our analysis on SDN architecture and performance at the OpenStack Summit in Atlanta.  We’ll continue to share our experiences in this and other areas of OpenStack.

 

How Do You Run OpenStack Securely?

Ok, we’re back to Symantec and security – naturally this is a critical criterion for us.  Security starts at the physical infrastructure, all the way through the services, and throughout operations.  It’s a very broad subject.  The OpenStack community is making the right moves towards ensuring the OpenStack software is secure, and we’ll get more involved in these efforts over time.  But a lot of your security will come from how you deploy, configure, and operate it.  I presented some of our thoughts on secure operation of Keystone at the Atlanta Summit.  I posed a lot of questions, and left some of them unanswered, TBD.  We’ll answer those questions in future postings, as well as discussing new questions around other aspects of OpenStack and security.

 

What Else Do You Need?

OpenStack provides a great base of IaaS services to solve our product teams’ infrastructure needs, and we’re going to use every bit of existing functionality that we can.  But there are features we need that don’t exist yet – we’ll be working with you all to build and contribute it.  In fact, there are some areas where we’ve found we need entirely new services and have started to build them from scratch such as batch processing, stream processing, and operational monitoring that our customers will use as OpenStack compatible services.

MagnetoDB is another one of these new services that Symantec is investing heavily in, a fully open source NoSQL Database as a Service for OpenStack.  The goal is to help our users to move most of their data from vertically scaling databases like Oracle and SQL Server to a managed service that scales horizontally for high throughput, capacity, and availability.  Look for MagnetoDB to be the subject of my next post.

 

We’re looking forward to continuing the discussion.  Stay tuned for more…

 

Keith Newstadt

Symantec Cloud Platform Engineering

Follow: @knewstadt