[Back to LISA '05 Call for Papers]

Want to write a refereed paper for LISA or give an invited talk, but don't have an idea for a topic? Here's a list of open questions and research areas that the LISA '05 organizers and a few colleagues created as a good starting place. We're especially interested in papers and talks that address the following topics, but we welcome proposals about all new and interesting work. Please see the refereed papers and invited talks sections of the Call for Papers for more details.

• Are all our sites really that different? Are we missing a more standardized methodology? Or are we just all rugged individualists who are causing our own problems (e.g., the account management problem)?
• Can you architect an infrastructure that won't be obsolete in 3+ years?
• Case studies: How do we move from reactive to proactive?
• Case studies: Sometimes we need to see the overall, integrated result, instead of yet another new tool to do the same thing some older tool already does
• Designing machine rooms for the next 3+ years
• DIY-IT (the "IT Garage Trend")
• Dynamic building of coalitions/collaborative environments
• Exploring collaborative tools and "social software"
• Exploring P2P, VoIP, and XML
• How can we design systems that fail-safe by default?
• How do people manage their personal email?
• How do you "manage" your manager?
• How do we train sysadmins to solve problems
• Improving tools for diagnosing problems with systems: Why can't our systems do a better job of explaining what's wrong with them?
• Information sharing: Need to know vs. publish and flag
• Innovative ways to exploit ticket systems
• Intel has suggested it's hit the limit with a 4Ghz processor; how do we deal with this cap?
• It's 2005: Why do we still have computer viruses?
• Managing content and collaborative systems for our customers: We are increasingly being asked to create and manage systems for our users to then expand and "grow." Consider the business use of blogs, wikis, and other systems that can have open-ended growth and change at the hands of "non-professional syadmins" (i.e., users). Do these have special challenges for sysadmins?
• Metrics
• Outsourcing/offshoring system administration: Is it possible?
• People management: More and more "classic sysadmins" are being pushed into first- and second-level management. How do we teach these high-geek people to manage people without growing pointy hair? What should sysadmins be learning to prepare for the day when they wake up managing 3–10 rugged individualist sysadmins (that were likely their peers the day before)?
• Real system administration tools: Why does every sysadmin/site have to roll its own tools?
• Scaling: How do you deal with the next 2x in storage, backup, networking, address space, database, productivity, etc.?
• Scripting languages (stuff around them)
• Selling sysadmin to management: What is your personal ROI? How do we measure it?
• Spam
• The "scaling problem": How do we scale (share) successful tools and processes, content creation and system extensions, and sysadmin experience?
• Tools for understanding information flows in networks and systems
• Virtualization: Benefit or bane?
• What can big sites learn from small sites?
• What have/haven't we learned about picking defaults?
• XML usage for configuration (management)
• Zero administration systems: Why not?