BABAR Computing System Home
Page
The BABAR Computing
System includes the following elements: online (data
acquisition, control, monitoring); offline (CPU,
network, storage); software
environment in which the computing work is done (including
code development/management, user concerns, distributed computing
issues, etc.); and the collaboration code itself
(on-line, reconstruction, simulation, etc.). This page
is an entry point for more information about this system.
News items:
26 Oct 2007 - See http://babar-hn.slac.stanford.edu:5090/HyperNews/get/physAnal/3697.html
for the announcement of the PR R22d-v04 datasets covering run1 through
run6. The reprocessing of the three week end of run6 period has
started. PC has completed its processing for runs 75550 to 76177
and ER has recently started on this range of runs excluding the SVT
reduced window test runs. The remainder awaits the new code for
processing the SVT reduced window data. Looking at R22d
skim production monitor you will notice that skim production has
recently had some difficulties. There are several causes: a problem
with merging certain streams for modes 1235 and 1237 that results in
exceeding the memory limits of the batch nodes, slower response from
the bbrora DB because of the huge number of entries, HPSS slowness
possible connected to the switch to faster machines (balis+fells) for
production that may have caused some choking of HPSS. All of
these problems are currently being attacked.
28 Sept 2007 - See http://babar-hn.slac.stanford.edu:5090/HyperNews/get/physAnal/3666.html
for the announcement of the recent PR R22d-v03 tag, there should be a
tag very soon for the R22d SP generics. Almost all of the
run6 generics have been skimmed with the exception of the SP
corresponding to the reduced SVT window period. About 45%
of each of the run1->run5 run cycle SP generics have been skimmed.
See http://babar-hn.slac.stanford.edu:5090/HyperNews/get/DQG/1502/1.html
for the R22c-v05 tag. Other news, the blackbox machines are up and
running. Skim jobs controlled by TM2 are currently being run on the
blackbox machines. A period of adjustments will be required to make
maximum use of the new resources. Lastly, the noma and tori
replacements are now believed to be in the racks and they should become
available for use later next week. Progress is being made on ramping up
skim production at GridKa, RAL/Manchester, Padova and we expect soon to
have contributions from in2p3.
14 Sept 2007 - The status of the R22d skims is continuously updated
in the tables shown on the
R22d
skim production monitor. In summary, all the run1-run5 m2m PR has
been skimmed except for 62
collections recently approved by the DQG and ~10 collections that crash
SkimMiniApp. The run 6 PR is
at ~72% completion and the remainder upto the beginning of the reduced
SVT window data should be
skimmed within the the following ~4 days. 30% --> 40% of the run1
through run4 generics have been skimmed
and >20% of run5 generics have been skimmed. The SP generics for
run6 January through June have
been skimmed. As of 13 Sep. 2007 the priorities have been adjusted to
be PR, SP signal and then
SP generics. Production of SP signal skims should start after the
processing of new PR has completed
(~4 days) and will last upto two weeks. At the current rate,
R22d skim production should complete near
mid-October.
1 Jun 2005 - Added link to the page
detailing the availablity
of datasets at Tier A sites
28 Jun 2004 - A Run 5 planning
page has been added
25 May 2004 - A new page
summarizing the "current" releases is now available.
15 May 2004 - The new WebSRT
system is being used for tag management
13 May 2004 - IOStreams Migration
page added
Older items
Fast links:
- Releases (Release
Summary, Lettered
builds)
- Online
- Offline (Reco,
Simulation,
QA,
Data
Quality, Software
Admin, WebSRT)
- Operations
(Production, PromptReco, CondDB,
Skimming Shift Signup)
- Distributed
(Tier A
data, Bookkeeping,
PromptReco,
Databases)
- Tools
(SRT, GUIs)
- Web
(Information, Authorship
guidelines, Tools,
Web
Usage Statistics)
- Misc (HyperNews,
Training,
New
Users, meeting
calendar, absence)
- Monitoring
(Ganglia,
Old
System Monitoring)
- Documentation
(Offline User's
Workbook, Intro to CM2)
- Task
Manager Version 2
Old news items (newest first)
Getting Help
Pointers to the Major Systems:
Plans, Designs, Organization, and Communication
User and Programmer Information
General Information about Computing
Stale Items
This page is maintained by Homer
Neal
This page was generated by the BaBar web unwrapping script. Original
page was:
http://www.slac.stanford.edu/BFROOT/www/Computing/index.html
|
