1 / 6

Xrootd Back To The Future

Xrootd Back To The Future. Andrew Hanushevsky SLAC National Accelerator Laboratory. WLCG Data/Storage TEG Meeting January 24-25, 2011 NIKHEF, Amsterdam, Netherlands. The Project Vision. Provide data a ccess in a manner which is Affordable

dima
Télécharger la présentation

Xrootd Back To The Future

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. XrootdBack To The Future Andrew Hanushevsky SLAC National Accelerator Laboratory WLCG Data/Storage TEG Meeting January 24-25, 2011 NIKHEF, Amsterdam, Netherlands

  2. The Project Vision • Provide data access in a manner which is • Affordable • Minimize hardware, software and human requirements • Highly performant • Minimize use of CPU, disk, memory and network resources • Robust • Minimize failure modes and bottlenecks • Secure • Support widely used authentication & authorization models • Scalable • Open ended scale up and scale out technology • Supportable • Collaborative model with a straight-forward architecture Storage/Data TEG Face To Face

  3. Vision & Experiments • The vision dovetails with experimental concerns • This is should be expected when one considers that… • Xrootd was developed from the start with • Experimental involvement during the development process • It continues to this day • The project is driven by experimental feedback • E.G. recent additions to the monitoring feature set driven by CMS • The project is well aware of the bottom line • Experiments don’t care how analysis is done as long as • The analysis gets done in a timely and efficient manner • I.E. it’s all about high energy physics not the computer science Storage/Data TEG Face To Face

  4. Vision & The Future • Xrootd is an extremely well hardened system • In production use for over 10 years in many environments • The number of new bugs is relatively small • Maintenance effort is very manageable • Our primary goal is to evolve what we have • To meet evolving experimental operating modes • The architecture is well positioned to do that • The plug-in model has proven to be a god-send • The vision and platform form a sound foundation • This serves us well to meet future requirements Storage/Data TEG Face To Face

  5. Some Things We See Comming • Solidifying the federated storage model • Make it even easier with a far more predictable outcome • More monitoring • The right information at the right time to the right places • More storage management features • Simple multi-tiered storage • We already do it but the documentation is sparse • Automatic storage rebalancing • Very sensitive to what underlying file system is used • Others as they bubble up in importance Storage/Data TEG Face To Face

  6. Conclusion • The next 3 years will be exciting • Nothing is really caste in concrete • The project is positioned to be highly adaptable • In a way that does not jeopardize existing production • Cluster IP is uniquely applicable elsewhere • To new and evolving technology and protocols • The architecture and software can work in many areas • The collaboration is expanding • Currently CERN, Duke, JINR, SLAC, and UCSD Storage/Data TEG Face To Face

More Related