1 / 11

DCache Basics

DCache Basics. Alessandro Usai, Sergio Maffioletti Grid Group CSCS. DCache General Architecture. DCache is flexible. At the core of DCache are subservices (cells) and domains (JVM) sub-services can run in different locations. General comments.

Télécharger la présentation

DCache Basics

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. DCache Basics Alessandro Usai, Sergio Maffioletti Grid Group CSCS

  2. DCache General Architecture

  3. DCache is flexible • At the core of DCache are subservices (cells) and domains (JVM) • sub-services can run in different locations

  4. General comments • It is all Java based i.e. platform independent (in principle) • It is highly configurable (good) but complex (bad)…the documentation could (needs to) be improved, the logs are obscure (if not absent, unless you explicitly increase the debug level) • Not easy to maintain upgrades • Yaim component not powerful enough (i.e. you do it by hand)

  5. DCache@CSCS

  6. In a nutshell • Pnfs needs to be mounted by the pool and admin nodes (they need to be authorized) • Postgres configured to grant the admin node • Thumpers installed by coping /opt/d-cache from the admin node and changing the configuration where required. Data and Control dirs are placed under /data1 and pointed to via sym links • CAs and grid-mapfile regular update from CE (via cronjob)

  7. Our Pools Configuration • Read only and write only pools (to balance the load) • p2p enabled (files are replicated for load reasons, only cache copies which get deleted if needed) • Two movers queues: wan (gridftp) and default (dcap) • 10 concurrent transfers at a time • se02-lcg with dteam+ops,cms,atlas and se03-lcg with dteam+ops,cms,atlas,lhcb,hone,dech

  8. Some problems we saw • Logging obscure (difficult to troubleshoot) • Domains duplication (as a result we got on board default configuration and a performance hit) • Timeouts (a known problem still under investigation within the community) • Various authentication problems due to misconfigurations (e.g. wrong VO tags)

  9. How are we doing just now?

  10. Tips and Advices • Recommended: increase the logging level • Install srmwatch (beware of the security holes though) • Get acquainted with the ssh interactive shell on the admin node (ssh -l admin -c blowfish -p 22223 localhost) • The dcache admin web page is very useful (http://storage01-lcg.projects.cscs.ch:2288/) • In case off problems user-forum@dCache.ORG

  11. References • http://twiki.cscs.ch/twiki/bin/view/LCGTier2/CSCSDCACHE_ausai • http://www.dcache.org/ • http://www.dcache.org/manuals/workshop2007/ndgf.shtml

More Related