Use of Condor on the Open Science Grid

Use of Condor on the Open Science Grid Chris Green, OSG User Group / FNAL Condor Week, April 30 2008 April 30, 2008 Condor Week Chris Green OSG User Group / FNAL

What is OSG? Links • Collection of mostly US-based • OSG home page. scientific / academic sites sharing • VORS resource map information. computing and storage resources • and VDT (Virtual Data Toolkit) home page. via common software stack. • Current use of OSG. • Job submission and management based around Globus / Condor. G. • "Virtual Organizations" (VOs): trust point for authorization; role-based personalities. • Works with multiple underlying batch systems (Condor, PBS family, LSF, SGE). April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 1

OSG facts and figures • 83 registered computing resources. • 30 registered VOs. • Usage breakdown for 2008/04/19 – 2008/04/25: April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 2

Survey of Condor use on OSG • Out of the box: «Condor. G for inter-site job transfer via Globus/GRAM: GT 2 submissions via Condor. G still (by far) the most common method of grid job submission on OSG. «Task scheduling for site health monitoring. «One of several batch systems supported on OSG. «"Managed. Fork" job management. April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 3

Survey of Condor use on OSG • External projects «Glidein / WMS: "pilot" job submission and management. «Fermi. Grid: job forwarding, "campus grid" management. «OSGMM / Re. SS: job forwarding and attribute-based matchmaking across multiple OSG sites. «"condorview: " enhanced job monitoring and control – not the web-based statistics client of the same name. «Complex workflows (eg LIGO: Pegasus/DAGMAN). «Gratia: accounting system leverages features of condor where available: condor_history, PER_JOB_HISTORY_DIR, DN. April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 4

More detail: Glidein/WMS • Workload Management System (Igor Sfiligoi, FNAL) uses Condor Glideins -- startd submitted as a grid job ("pilot") makes remote batch nodes look like local ones. • Two main components: «One or more glidein factories: manage available grid sites and submit pilot jobs. «One or more VO frontends: receive payload submissions from users for distribution to sites. • Pilots receive user payloads as distributed by VO frontends. April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 5

More detail: Glidein/WMS April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 6

More detail: Glidein/WMS • Uses GCB for firewall / NAT management. • Intra-VO priority management. • Works with gl. Exec: application running on worker nodes which handles authorization and UID mapping for payloads – per user accountability to the site. • Unaffected by grid site batch manager choice. • V 1. 0 released Dec. '07; v 1. 1 Jan'08. • In use by: CDF; Minos (FNAL); being commissioned for CMS. April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 7

More detail: "condorview" • Michael Thomas, Caltech. • Graphical tool for browsing and managing a condor queue. • Hooks to vacate and kill jobs. • Hooks to ssh into job directory on worker node and print out process tree. • Uses condor_q, condor_config_val, and condor_fetchlog. April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 8

More detail: condorview April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 9

More detail: condorview April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 10

Concluding statements • Condor essential to the OSG. • Condor use underpins connectivity of sites within the OSG. • Close ties: Miron is OSG PI; VDT team at Wisconsin; new Condor features often a result of OSG needs. • Widely used on OSG; many novel uses of and applications building on Condor features. • More details in later talks! April 30, 2008 Condor Week Chris Green OSG User Group / FNAL 11