CMS: Difference between revisions

From ReddNet
Jump to navigation Jump to search
No edit summary
 
(24 intermediate revisions by 2 users not shown)
Line 1: Line 1:
==Goals==
=== UNDER CONSTRUCTION ===
*L-Store plugin for root, so that rootio can read (write?) L-Store files
*This page is for CMS Users interested in using REDDnet - Still under construction, but details will be added soon!
*CMS then uses REDDnet as temporary storage for user analysis (Tier 3 and below)
*Plugins for CMS code have been developed so users can easily access REDDnet:
*Other CMS Applications possible, begin with above.
**Command line tools for file upload and download into REDDnet: LoRS, Lstore
 
**SRM and Phedex: GridFTP plugin  
==Benchmarks==
**CMSSW: Root plugin
*IBP --> CMSSW streaming tests
*[https://promise.cs.utk.edu/lodn/cgi-bin/lodn_published.cgi?subdir=/cms_reddnet LoDN exnode server]
**CMSSW_1_2_0
*We have a small group of Tier 3 users testing submitting CRAB jobs to OSG sites with data on REDDnet.
**ROOT/L plugin
***extends TFile Class
**input: 1.5GB root file, 100 events
**Naive implementation
***relays block requests directly from root
***no blocksize specified in root
***no caching or read-ahead
* Note
** PING measures latency
** IPERF & File Download measure bandwidth
* RUNNING ON VAMPIRE, 2GHz CPU:
 
{| border="1"
|+align="bottom" style="color:#e76700;"|''Vanderbilt Campus has a 0.60 Gbit/s connection to outside''
 
! rowspan="2"|Data Source
! rowspan="2"|# Depots
(stripe)
! rowspan="2"|URL
! rowspan="2"|PING
time (ms)
! colspan="2"|IPERF
! rowspan="2"|File (lors)
Download
! rowspan="2"|CMSSW Job
100 events (mins)
|-
! in !! out
|-
|local gpfs || 0 || /gpfs2/ || || || ||
| style="color:#e76700;"|''11''
|-
|Vanderbilt REDDnet Depots || 10 || vudepot1.accre.vanderbilt.edu || 0.16 || || || ||  1.0
|-
|across campus IBP Depot || 1 || vpac12.phy.vanderbilt.edu || 0.46 || || || || 3.5
|-
|rowspan="5" | Remote IBP Depots
|rowspan="5" | 5
|ounce.cs.utk.edu || 13 || ||
|rowspan="5" |
|rowspan="5" | 20
|-
|pound.cs.utk.edu ||  13 || ||
|-
|acre.cs.utk.edu ||  13 || ||
|-
|umich-depot01.ultralight.org || 83 || || 
|-
|ibp.its.uiowa.edu || 35  || ||
|}
 
* RUNNING AT CALTECH, 2.4GHz Opteron:
 
{| border="1"
|+align="bottom" style="color:#e76700;"|''Caltech has 10 Gbit/s connection to Ultralight''
 
! rowspan="2"|Data Source
! rowspan="2"|# Depots
(stripe)
! rowspan="2"|URL
! rowspan="2"|PING
time (ms)
! colspan="2"|IPERF
! colspan="2"|1.5GB File Download
! rowspan="2"|CMSSW Job
100 events (mins)
|-
! in !! out !! time (m:s) !! rate (Mbits/s)
|-
|local disk || 0 || /dev/hda3 || || || || || || 1.0
|-
|Vanderbilt REDDnet depots || 10 || vudepot1.accre.vanderbilt.edu || 78|| || || ||
| style="color:#e76700;"|''100''
|-
|U Mich REDDnet depots || 2 || umich-depot01.ultralight.org || 63 || || || 4: 0
| style="color:#e76700;"|''51'' || 33
|-
|Caltech REDDnet depots || 2 || caltech-depot01.ultralight.org || 0.60|| || || 0: 16 || 770 || 1.0
|-
|UT Knoxville IBP depot || 1 || acre.cs.utk.edu || 93|| || || || ||
|-
|U Iowa IBP depots || 1 || ibp.its.uiowa.edu|| 70|| || || || ||
|}
 
==Current Work in Progress==
*Figure out how to get necessary code included in CMSSW
**Talk to Bill Tannenbaum, Phillipe Canal,...
**include L-Store code in CMSSW distribution so it is built on platforms correctly for use with rest of CMS software.
**that way no software for users to download themselves, no changing of configuratino scripts, etc.
**how test and validate before checking in?
**how to check code in?
*Figure out all issues needed to integrate with CMS data distribution model
**phEDeX, TFILE, DBS/DLS,...
*Switch Root plugin to use L-Store version of libxio
 
==Demos==
 
===Demos at March 2007 OSG Grid Meeting (UC San Diego)===
Can use Vladimir's or Dmitri's analysis for all of below.
 
====Interactive Root====
*First upload a file to Depots spread across the WAN, use LORSView to show where and how they go.
*Then read it back in root, show it works.
*Mainly an introduction to the issues.
 
====100 Node ACCRE Run====
*each reads its own file from WAN set of depots.
*show speed versus local copies of file (data tethered analysis).
 
====100 CPU Grid Job====
*similar to ACCRE Run, each job reads its own file from WAN depots.
*jobs are distributed accross open science grid sites
*demonstrates complete lack of data tether.
 
====To Do To Get Ready====
*Run all of the above many times before actual demo!
*Get LORSview upload working
*Figure out how to submit 100 CPU Grid Job.
*Want to run all 100 ACCRE jobs simultaneously?  Need to work with ACCRE on that...
 
===Get Rick Cavanaugh to run his analysis===
*need most of the stuff needed for "Summer 2007 demo" but maybe not all fully in place.
*he runs himself.
*work with him so he understands full functionality possible.
*work with him to develop ideas for better implementing Summer 2007 demo
**what docs are needed
**best approach to getting users using it
**etc.
 
===Summer 2007 Demo===
*A "scratch space" demo for CMS users.
*Use deployed REDDnet resources which should become available June 2007
*Load REDDnet with "hot" data files, convince a few users to try them out
*Must have L-Store code fully integrated with CMS software
 
==General Testing==
===verify ROOT/L works===
*   package up plugin for CMS L-Store test community
*    gain experience via benchmarking
*   finalize API (add write access?)
*   checkin plugin to root cvs
**        it will take a while for this ROOT addition to propagate into CMSSW
*    explore CMSSW procedures for checkin of LORS and/or LSTORE
**        it will take months for L-Store to be available for check-in
 
===increasing level of stress tests:===
(validate and benchmark)
 
====do various combinations of the following:====
* single jobs vs simultaneous jobs
* many jobs: at one cluster vs across the grid
* simultaneous jobs hitting same ibp depot accessing one or many files
* simultaneous jobs hitting same file at one depot or striped across many depots
 
===Also need to benchmark and profile various types of jobs:===
* I/O intensive skims
* CPU intensive jobs
* show benchmarks/demonstrate which jobs work well with L-Store and which jobs won't work well (if any). Have thorough benchmarks for the worst-case scenario.
*gather numbers to discuss impact on bandwidth as L-Store usage explodes.
*will people feel more free to do unnecessary computations?
 
== assemble interactive analysis demos:==
* host variety of interesting datasets
** need to identify these datasets
* make a wiki
**        with instructions
**        links to necessary data catalogs
***            L-Store
***            DBS/DLS
* gather visually interesting ROOT Macros
**        event/detector displays
**        histograms/results
* any FW-lite tools (even development versions) to try
 
==assist user-analysis batch production:==
* identify and host a wide variety of datasets
**        calibration datasets
**        various backgrounds, pileup
**        variety of signal samples
* populate catalogs to find datasets
* web tools to assist this
**        how to find datasets
**        how to upload results
**        how to register results in catalogs
**        how to coordinate with L-Store and DBS/DLS
 
==provide info on joining CMS/L (via L-Store?  via REDDnet?  ultralight?)==
*    how to add an ibp depot
 
 
==Long Term Needs==
*L-store version of libxio
*Stable, production REDDnet depot deployment
**'''including request tracker support!'''
*L-store software fully integrated with CMS software, and being distributed.
**'''this means need source code for L-Store version of libxio - check into CMS distribution'''
 
==SRM interface==
*in principle, this is important for CMS usage. 
*'''Need to get new support person on board with this and up to speed.'''

Latest revision as of 18:04, 10 April 2008

UNDER CONSTRUCTION

  • This page is for CMS Users interested in using REDDnet - Still under construction, but details will be added soon!
  • Plugins for CMS code have been developed so users can easily access REDDnet:
    • Command line tools for file upload and download into REDDnet: LoRS, Lstore
    • SRM and Phedex: GridFTP plugin
    • CMSSW: Root plugin
  • LoDN exnode server
  • We have a small group of Tier 3 users testing submitting CRAB jobs to OSG sites with data on REDDnet.