Systems management from the cloud
Monthly Archives: January 2011
January 18, 2011Posted by on
Many thanks Erick Schonfeld and the team at TechCrunch for the great writeup on what we’re up to:
I think the article hits on the core difference in what we’re trying to accomplish – “Replace million-dollar deployments that take months with a five minute download that can have a machine being managed from the cloud in five minutes.”
We hope to deliver on this vision. If you’re interested in trying out the product (of course, an early version) just sign up on our website for updates and we’ll send over a login when we’re ready to open things up for early access.
January 15, 2011Posted by on
Enterprise IT infrastructure management is due for a change. Around the world, social computing and elegantly simple software have combined to revolutionize consumer products. Yet enterprise IT still relies on plodding, hugely over-priced solutions. It’s no wonder most sysadmins rely on their own tools or manual processes to manage their systems – whether those are physical machines, machines sitting in their private clouds on VMware, or even machines sitting in public cloud systems like Amazon EC2 or Rackspace.
Technology analysts peg worldwide annual sales of systems management and datacenter automation software at over $14 billion. Most of the sector remains largely dependent on traditional enterprise IT solutions – solutions that require systems administrators and their host enterprises to deploy large, complicated software stacks. Then it’s time for everyone to learn new languages and configuration processes just to get older, previously working applications redeployed after this “transformation”.
There is almost every form of product available in this market. The high-end has a handful of players; analysts have even taken to calling them the “Big 4” – Computer Associates, BMC, IBM, and HP. Open-source providers such as Puppet Labs and Opscode sit at the other end of the spectrum and are trying to own the open-source charge. And in the middle, there is the Microsoft Systems Management Server and InTune, focused on Windows as a specific platform to be managed.
Every one of these products is still a big software package that is complicated to deploy. Between software deployment consultants, training, new infrastructure (databases, etc.), new languages and new processes, deploying this software to manage even a small number of servers is a huge and complicated task. Here are some other specific things that you’re probably aware of when you’re dealing with these solutions:
- Expensive: If you have ever tried to work through a sales cycle with one of the traditional vendors in this space, the biggest shock you will get will be how much their stuff will cost you. At anywhere from $1,000 to $15,000 per server plus a 25% maintenance fee, you’re paying a lot of money. How can you justify the costs for this software as you spin up multiple virtual machines on your VMware datacenter or pay by the hour over at Amazon EC2?
- Difficult to Deploy: The next shock you get working with an incumbent in this market is how complicated it is to deploy these products on your infrastructure. Remember, each of the vendors in this space makes a lot of its revenue from implementation services. And the open-source software vendors make all of their money this way. Once you pay them for the licenses you will most probably end up with an army of consultants working to “re-engineer” your processes to match their product, or “customize” their product to match your processes. You’re stuck either way you go.
- Inflexible: Customers are increasingly running a diverse set of infrastructure—old physical machines, private-clouds running on VMware, public-cloud machines on Amazon EC2 and Rackspace, etc. Each one of the existing software suites is trying to retrofit these new systems into their creaky old framework. With so many customers are starting to re-think the way they use compute cycles. Why be stuck with an old, on-premise model of management that haphazardly bolts on these new pockets of compute power? Instead, they might prefer to go with a clean slate, something designed from the ground-up to address the current state of computing.
- Anywhere Access: Sitting at a bar and trying to figure out what is going on with your IT systems? Hanging out with your iPad on the couch? Don’t hold your breath waiting for one of the “Big 4” systems to give you access to your systems anytime and anywhere you want them.
We could go on and on about the problems with existing solutions, but we often find that customers have their own long list of gripes. We feel that the old model is broken, and that there has to be a better way to help people manage their systems.
At ScaleXtreme, we’re working on something better. Since we’re not announcing our specific product just yet, we’d like to share some core principles that are guiding our product design:
- Simple: Our goal is to make something that ordinary IT administrators can deploy and run in a few minutes. That will be a big improvement on the complicated product model forced onto customers that requires an over-the-top learning curve. Most customers actually have little or no real automation deployed to help manage their systems, primarily due to the cost and complexity of deploying these solutions. We’d like to help these folks.
- Scalable: We are a cloud-based system, and we’re designing our system to handle a single server or hundreds of thousands. If you’ve ever had to hire an IT pro just to manage the database of your systems management product, you might appreciate this. But being scalable goes beyond just being in the cloud. We will support a surprising number of platforms. Our service will tackle the most heterogeneous and complicated enterprise IT systems out there. And you can buy as many licenses as you need right from our website, whether that’s one, one hundred, or one thousand. We aim to serve everyone from the smallest startup companies to the largest multi-national corporations.
- Social: Building in the cloud from the beginning allows us to bring the best of collaboration and content to systems management in a way that hasn’t been done before. Systems management, just like any other task in large enterprises, is a team effort. It is surprising how “un-social” most products in this space have been, especially when it comes to leveraging the knowledge and community available on the internet, and not integrating this directly into their products. We’re excited about this topic and will write more about it in future blog posts.
Thank you for all the positive feedback and support so far. If you are interested in being part of our early-access program, please sign up on our website and we’ll try our best to accommodate you. We look forward to delivering a great product to market that will make a difference in the scalable management of systems out there, wherever they may be.