Make High Availability Work for You

The first Open High Availability (HA) Cluster Summit will be held on May 31st, 2009 in San Francisco, California. Please join us for the latest updates on Open HA Cluster technologies as well as key insights from HA Clusters community members, technologists, and users of High Availability and Business Continuity software.

This is a free event.

Venue Open HA Cluster Summit Sunday, May 31st, 2009 Marriott Hotel (Next to Moscone Convention Center) 55 Fourth Street, San Francisco, CA 94103 USA http://www.marriott.com/hotels/travel/sfodt-san-francisco-marriott

Twitter If you would like to tweet about the Summit, or would like to follow it on twitter, use the hashtag '#ohac'. Please use twitter to ask online questions, as there is no live chat for the session.

Live Stream Join us for a live streaming broadcast if you cannot attend in person. Morning sessions and both afternoon tracks would be broadcast live.

Good Things To Know Wireless internet connections will only be available to presenters, so you may choose to leave your laptop at home. Best off-site parking is at 5th and Mission. To arrive at the Summit location, it is best to enter through the hotel front entrance. Go slightly to the right and go down two sets of escalators to the Yerba Buena Ballroom, Salon number 4. Look for signage.

Back to top

Registration And Contact Information

Online self-registration is now closed. Please email ha-cluster-summit-info AT opensolaris DOT org for late registration or more information. There could be a limited number of walk-in registrations available at the venue. If you would like to follow the Summit on the internet, there will be a live webcast of the talks including both technical tracks in the afternoon.
If you have any questions about the summit, please email ha-cluster-summit-info AT opensolaris DOT org.

Back to top

Schedule

Download Agenda pdf

Back to top

Morning Speakers

Keynote: "The Network is the Cluster" High availability is central to wide range applicability of clusters; no applications like the alternative, non-availability. Here, the network is key given it provides the coupling between computers to form the cluster. The network is basically a set of highly available computers (switches) that, besides forwarding packet, have good connectivity to all. This talk looks at exploiting extensibility of switch software to provide basic code services to support high availability. Dr. David Cheriton is an industry expert and Professor and Head of the Distributed Systems Group, which is the research group in the Departments of Computer Science and Electrical Engineering at Stanford University. He co-founded Granite Systems and Kealia with Andy Bechtolsheim (Sun co-founder). He was instrumental in helping Stanford students Sergey Brin and Larry Page fund Google. In his "spare time", Dr. Cheriton is Chief Scientist at Arista Networks and serves on the board of Aster Data Systems.

Project Colorado: A minimal and modular High Availability Cluster for OpenSolaris Project Colorado is a next-generation high availability cluster for OpenSolaris with reduced hardware requirements and modular software for easy customization. Colorado allows you to run an HA cluster on OpenSolaris without shared storage, redundant NICs, complicated cabling, or quorum device by taking advantage of COMSTAR iSCSI and Crossbow VNICs. Includes a live demonstration. Nicholas Solter is the lead author of OpenSolaris Bible (Wiley, 2009), a 1000-page reference and tutorial on OpenSolaris. Nick has worked at Sun Microsystems for more than eight years in the areas of high availability and distributed systems. In his work on the Solaris Cluster product, he has implemented clustering support for core Solaris features such as Zones and SMF. He was the technical lead in open sourcing the Solaris Cluster product and is currently leading Project Colorado.

Back to top

Guest Panelists

High Availability: Today and Tomorrow

	Tasso Argyros,  Aster Data Systems Bio:  Tasso (or Tassos) is Chief Technology Officer and Co-founder, of  Aster Data Systems, San Carlos, CA.  Tasso took a leave of absence from the Ph.D. program at Stanford University to co-found Aster Data Systems. As a Ph.D. student, he was a member of the Distributed Systems Group and investigated extreme-scale cluster architectures for fast, parallel data processing across large farms of commodity servers. During his tenure at Stanford, Tasso was a SAP Graduate Fellow and earned a Master's degree in Computer Science. Before studying at Stanford, he conducted research in data mining for temporal and web databases at the Technical University of Athens.
	Mark Callaghan, Google Mark has led the MySQL Engineering team at Google for 2.5 years. His current focus is on making MySQL perfect. He has more than 13 years of database internals experience at Informix, Oracle and now Google. He has an M.S. in Computer Science from the University of Wisconsin-Madison. He maintains a large MySQL patch at code.google.com and a smaller patch at launchpad.net. His blog is at http://mysqlha.blogspot.com.
	Sreeram Duvur, Sun Microsystems Bio: Sreeram Duvur is a Principal Engineer at Sun Microsystems who has been working on Enterprise and Telecommunications middleware and the Java Platform for over a decade. He is currently the architect for Project SailFin, the leading open source Communications Server.  He previously led the GlassFish Enterprise Server and Java HotSpot VM engineering teams at Sun. He holds an MSCS from University of Oregon and an MS in Management Science from Stanford University. His interests include distributed systems, high availability, data mining and computer systems performance.
	Dr. Ira Pramanick, Google Ira Pramanick is a Staff Software Engineer and works in the Search Appliance (Enterprise) group since she joined Google a year ago.  Prior to that, she worked for 9 years at Sun Microsystems in the Solaris Cluster group.  She has also worked at Silicon Graphics, IBM and the University of Alabama in Huntsville. Ira is on the Editorial Board of the Cluster Computing journal and is IEEE-TCSC's coordinator for High Availability.  She is a holder of 4 US patents.
	Hartmut Streppel, Sun Microsystems Hartmut has worked for more than 25 years in the IT industry.  He is now a Principal Field Technologist with Sun Microsystems, Germany, and a well known Solaris Cluster specialist in the field.  Before joining Sun Microsystems in 1999, he got first hand experience with clustering solutions at DEC.  Hartmut has also worked for the Open Software Foundation on distributed management.

Back to top

Technical Tracks Descriptions

Title: Glassfish Clustering for Telecommunications
As subscriber population rapidly increases carriers are increasingly replacing legacy switches with open standards based systems using IP, COTS hardware and middleware. Java EE platforms are becoming popular as the services delivery platforms of choice because of the maturity, richness of available APIs and abundance of developers. In these platforms, scalability and high availability are paramount because outages can impact lot of paying customers. Horizontal scaling, fast failure detection, failover and failback, integration in to an existing low level clustering platform are some key challenges. This talk will look at how GlassFish Enterprise and Communications Server has been integrated into very large scale Telecom services delivery solutions.

Title: MySQL Cluster and Sun Cluster for Carrier-Grade Availability
This talk will examine the architecture of MySQL Cluster, a shared-nothing distributed database system. We will show how MySQL Cluster and Sun Cluster are complementary technologies, and describe why and how a large telco customer chose to deploy MySQL Cluster and Sun Cluster together in a single application.

Title: Running Open HA Cluster with VirtualBox
Developers often like to carry their whole development environment with them on a laptop. This session explains how to setup a system with OpenSolaris serving as host for at least two VirtualBox OpenSolaris guests and shows how to setup a two node Open HA Cluster with them by using technologies like Crossbow and COMSTAR. Such a system can be used to build, develop and test Open HA Cluster, which will get demonstrated live during the session.

Following the VirtualBox demo, there will also be a brief overview of how to use OHAC with other virtualization technologies like xVM and LDoms. Using guest domains as cluster nodes vs black boxes to be failed over, some considerations in using virtual devices with OHAC, are some of the key topics which will be discussed.

Title: Getting started with Open HA Cluster
This talk will tell all about the Open HA Cluster community and how community members can participate in it. Even if you are not a hard core engineer, feel free to attend the talk as there is something for everybody in the community.

Title: Developing Custom Application Data Services
Data Services (also called agents) are the glue between an application and the Open HA Cluster framework. Agents make applications highly available on Open HA Cluster. This session explains how to develop an agent using the various tools and API available in Open HA Cluster. This presentation will give an overview of the application characteristics to be able to run as failover or scalable services on OHAC, Resource Management Model, DSDL/SCHA API, SCDS builder and GDS coding template.

Title: Why Business Continuity (BC) matters, and How to Extend Open HA Clusters/Geographic Edition to Fit Your Enterprise BC Plan
This talk will give an overview of how Disaster Recovery fits into a Business Continuity plan. It will then consider how to extend Open HA Clusters/Geographic Edition to control custom replication modules, by developing a new module using the Script-Based Plugin, as a part of a Disaster Recovery solution.

Title: Testing OpenSolaris Cloud Clusters
This presentation discusses how to use the resources on http://opensolaris.org to test Open HA Clusters including, configuring and running Open HA Tests Suites, developing and testing cluster software in the OpenSolaris Test Farm and using the Sun Open Cloud Platform.

Title: OHAC and AVS to Provide High Available Services Using NON-shared Disks
This presentation will show how to use the OHAC-AVS project to provide an option for users to have High Available solution using NON-shared(cheap) disks. Totally based on OpenSource softwares: OpenSolaris, Open Cluster (Sun Cluster Express), and StorageTek Availability Suite (AVS).

Back to top

Community Marketplace

There will be a Community Marketplace in the evening following the technical sessions. This will be an informal venue to showcase your product, or if have an interesting demo, or simply a presentation of your ideas, please sign up below. We can provide a table and necessary power for your setup. If there are special requirements, email us.

Evening Event - Casino Night

We're hosting a Casino Night where everyone plays to win! Your favorite table games, including roulette and blackjack will be available to play. Come in and find some of the best odds this side of Las Vegas! Fun money will be provided and you can play to win some great prizes. Perfect time to relax, kick back and have a little fun. Light hors d'oeuvres and drinks will be served.

Back to top

Pictures, etc.

Full sized images

Back to top

Attendance List

Please add your details below:

Thank you for your interest in this event. Online self-registration is now closed. Please email ha-cluster-summit-info AT opensolaris DOT org for late registration or more information. There could be a limited number of walk-in registrations available at the venue. If you would like to follow the Summit on the internet, there will be a live webcast of the talks including both technical tracks in the afternoon.

Back to top