Oracle Solaris 11 Outlined
By firstname.lastname@example.org on Sep 20, 2010
And as I'm writing this, John Fowler and Mark Hurd are delivering even more. Sun's new T3 based systems, powered by the worlds first 16 core processor, and Oracle Solaris 11, the only completely virtualized OS in the industry giving you even more efficiency and availability in your data center.
Any guesses as to what I'm going to focus on here and and tell you a bit more about?? :-)
Oracle Solaris has been setting the bar for enterprise OS innovation, reliability, performance and scalability for nearly the last 2 decades. As the #1 Unix OS by Volume, Oracle Solaris powers over 50,000 customers and is the #1 deployment platform for the Oracle DB. And with industry innovations like ZFS, DTrace, Solaris Containers, and Trusted Extension, Oracle Solaris 10 has continued that tradition to become the platform for 10 of the Top 10 Banks and 10 of the Top 10 Telcos.
Today, we announced we're raising the bar again with Oracle Solaris 11. Comprising more than 2700 projects, over 400 new inventions, 20 million person hours of development work, and over 60 million hours of testing, we'll deliver Solaris 11 in Calendar Year 2011, with the very first release of Solaris 11 Express coming this CY (2010). Solaris 11 Express will be a fully production supported platform, designed for customers that want to deploy the latest technology as soon as possible, and the platform for the new Oracle Exadata database machine and the Oracle Exalogic Elastic Could.
But what's in 11 you ask? What makes it cool? Why do you care? Glad you asked!
Oracle Solaris 11 is expected to focus on 4 main areas of problems customers all face today, getting the most Availability, Efficiency, Security, and Peformance and Scale out of their Data Centers. There, that answer your question? OK, maybe not by itself. So here's the details:
Availability is more than ensuring your systems stay up and running, surviving hardware failure through OS feature like our Predictive Self Healing framework, HA cluster solutions or even just core OS stability. It means addressing planned downtime as well and unplanned downtime.
When I used to run my small data center, I spent more time dealing with patching and OS upgrades that just about anything else, and because I was stuck with older technologies like mainframes, Windows, and Netware, I was stuck with ugly patching processes and upgrades that were always my hope and pray moments; that once the systems completed their patch/upgrade actions that were out of my hands, they'd actually work when I was done. It was a high stress period, and one where I'd take my mission critical services off line, late at night and hope I could get them back by the time folks showed back up for work in the morning. In today's world for most customers, there are no late nights, business occurs 24/7/365 globally and maintenance windows are hard to come by and squeezed constantly to the smallest windows possible.
So, how does Solaris 11 help us fix this? Well one of the biggest ways is IPS. What's IPS? It's the Image Packaging System. A dependency aware, network aware packaging system that virtually eliminate patching errors (the Unix Systems 5 dim sum patching problem is gone!) and gives you safe upgrades by combing with ZFS as the root filesystem to provide you the ability to prepare your system for upgrade live, without taking down your production services, by preparing an alternate boot environment so your upgrade process becomes only a restart. And, because your existing environment is never over written, should anything go wrong you can simple roll back to your old environment. And, through this process we've eliminated 50% of the restarts needed with the original Solaris 10 release, and many of the common restarts for more complex patching actions. Not to mention the new Fast Reboot technology that brings simple restarts back up in seconds versus minutes.
Now, if I keep up this level of discussion for each one of these areas, we'll be here for a while so I'm going to go a bit faster now, and then write a few more over the remainder of this week to give the focus to these new areas.
Data centers today must also be Efficient, Cloud based deployments require complete virtualization of our infrastructure, from Compute, to Network and Storage and Oracle Solaris 11 is expected to deliver all of that for the first time anywhere fully integrated into the OS. Building on our Solaris Containers technology, we will integrated the already delivery project crossbow technology into Solaris 11. Giving you both the ability to completely virtualize your network, through virtualize NICs and switching (in a completely standards based way so your existing network management tools can see your network!) to achieve higher levels of control and observability into your virtualized application infrastructure.
But Efficiency in your data center is not just about virtualization, it's about your human capital as well. What tasks does your infrastructure require your administrators to deal with? By deploying virtualization, is it solving a problem or adding an additional layer to manage? How many OS's and Virtual layers do you have in your data center and how much time does your team spend managing all of it? And what happens when you upgrade your Applications and OS's and VM layers? Are all of these things designed to work together? Does you OS scale vertically to the largest systems in the world (hundreds of cores, 1000's of threads, TBs of memory) and on horizontally scaled industry standard systems? There is only 1 OS that does this today and in the expected future, Oracle Solaris.
And again, with Oracle Solaris 11 we will continue to provide application binary compatibility, guaranteed, so your applications stay up and run in your new OS. We'll give you a Solaris 10 branded zone with P2V and V2V tools to bring now just your apps forward, but your entire environments as well. We're also modernizing the OS so you can reach into broader pools of administrator talent to keep your costs down, by providing the GNU tools in the default path, updates the shell to bash, and giving you a modern GNOME desktop with all the cool wizzy things like compvis so you have a great desktop for development and administration.
Though you said you weren't going into great detail on everything? OK, I'll spend more time this week discussing many of the other innovations we're planning, around Security for sure (ZFS crypto anyone?) and the Performance and Scale you need to run todays systems, and most importantly tomorrows.
So, welcome to the start of the show. What a great way to kick off the week! Hope to see you here, at any of the Solaris sessions this week or around the halls and events at the show.