[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
JLab Hall D computing manpower estimates
Hi,
Attached is my estimates of JLab Hall D computing manpower needs, with hiring
timelines and discussion. This note complements my notes on JLab permanent and
term hardware support staff needs. As with my other notes, this hiring timeline
will be included in Elton's Cost/Manpower spreadsheet.
My JLab Hall D staffing estimates are complete except for the total number of
physicists needed (I estimate 6 dedicated to online/offline computing). Larry
Cardman has indicated that Hall D will have roughly the same number of
physicists as CLAS, leading us to include 13 physicists total in the
spreadsheet.
Sincerely,
Elliott
================================================================================
Help feed the hungry, one click at a time: http://www.thehungersite.com
"Science plays a major role in satisfying the minor needs of mankind"
================================================================================
JLab Hall D Computing Staff
---------------------------
E.Wolin
21-Mar-2001
Introduction
------------
In this note I estimate the number of JLab staff needed to help
develop and maintain Hall D computing hardware and software systems.
I further comment on management strategies for these efforts, mainly
from a JLab perspective. How JLab management and Hall D group
coordination coexist needs further discussion.
Scaling from CLAS is difficult in this case since Hall D needs a
sophisticated computing system and highly reliable software very early
in the experiment (e.g. portals and grids, Level 3 trigger,
simulation, etc.), and we are planning to implement a distributed
computing environment from the start. Further, there are large
uncertainties in the level of support we can expect from the
universities. In contrast, scaling from CLAS for hardware support
personnel was straightforward.
Below I assume JLab Hall D staff will develop and maintain the Hall D
online computing system, lead the development of the early Hall D core
software and computing infrastructure, and will share the remaining
work with university personnel. I further assume JLab Computer Center
staff will maintain the silos, offline farms, networks, software,
etc. needed to transfer Hall D data to the central storage facility
and to process part of the Hall D data.
In previous notes I performed similar analyses for detector and
hardware support staff, and discussed non-Hall D online computing
staff (DAQ, Computer Center, etc.), and Hall D online system
management staff.
Distributed environment
-----------------------
We plan to maximize participation by university personnel through use
of a portal-based grid computing environment. Here the entire
computing framework is based on internet accessible tools so that it
makes little difference where developers or users work, assuming they
have high-speed internet connections. The scientific community is
devoting a tremendous amount of manpower to making grid and portal
systems work, and we will build on these efforts.
Note that successful implementation of these models requires a large
up-front effort at JLab to develop the necessary infrastructure.
Management at JLab
------------------
Management of the Hall D online effort must be performed by JLab staff
due to the intimate connection between the online, detector hardware,
JLab environment, and personnel involved. I propose a single position
be created to manage the Hall D online, coordinate the Hall D online
group, and supervise all online personnel.
Coordination of the Hall D offline group may be performed by
university personnel, although supervision of JLab offline personnel
must be done by JLab staff. If the offline coordinator is not a JLab
employee JLab offline personnel could be supervised directly by the
Hall Leader or by a local JLab offline supervisor. Alternatively, the
online group manager position could be expanded to become the head of
JLab Hall D computing, covering both the full online and the local
offline effort.
Role of JLab Hall D personnel in online
---------------------------------------
The online group's responsibilities include DAQ hardware and software,
trigger system, computers and networks, Level 3 farm, slow controls,
databases, monitoring and alarm systems, event format, etc. This
group must work closely with the offline, electronics, and detector
groups. Staff recommendations below assume non-JLab personnel will
contribute significantly to the Hall D online effort.
Role of JLab Hall D personnel in offline
----------------------------------------
Although JLab will be one of many institutions working on Hall D
offline reconstruction and analysis, development of critical
components needed early in the experiment might best initially be led
by JLab staff. These include the grid/portal infrastructure, the
initial simulation, reconstruction, and calibration frameworks,
database schema, data formats, data distribution schemes, etc.
Staff recommendations below assume major contributions from non-JLab
staff to the Hall D offline effort.
Role of university personnel
----------------------------
So far few university groups in Hall D are primarily committed to
computing and software development, and it is difficult for groups
developing major detector components to also be responsible for major
pieces of the Hall D computing system.
Major university tasks include setting up simulation, analysis, and
PWA analysis farms. The Hall D calibration farm may also be set up by
university personnel although it will likely be located in the Hall D
counting house. University groups are expected to play a major role
in the development of simulation, calibration, reconstruction, and
analysis code. Finally, we expect significant university
contributions to the online effort, implying long-term deployment of
university personnel at JLab.
Role of JLab Computer Center personnel
--------------------------------------
As with CLAS, the Computer Center will need to maintain the data
transfer and storage infrastructure, networks, simulation and
reconstruction farms, etc. used by Hall D at JLab, as well as assist
with system management tasks in the counting house. We expect most or
all of the first pass reconstruction will be done at Jlab, but that
most of the simulation and analysis of Hall D data will be done at the
universities.
Hiring Timelines
----------------
Fiscal Year | 1 | 2 | 3 | 4 | 5 | 6 | 7
Hall D Staff (long-term)
------------------------
Physicist (online) | 0.5 | 1 | 2 | 3 | 3 | 3 | 3
Physicist (offline) | 0.5 | 1 | 2 | 3 | 3 | 3 | 3
Computer Professional | | | | 1 | 1 | 1 | 1
--------------------------------------------------------
Sub Total | 1 | 2 | 4 | 7 | 7 | 7 | 7
Jlab Staff (non-Hall D)
-----------------------
Computer Center | | | 1 | 1.5 | 2 | 3 | 3
--------------------------------------------------------
Sub Total | | | 1 | 1.5 | 2 | 3 | 3
--------------------------------------------------------
Total | 1 | 2 | 5 | 8.5 | 9 | 10 | 10
--------------------------------------------------------
begin:vcard
n:Wolin;Elliott
tel;fax:757-269-5800
tel;home:757-229-2724
tel;work:757-269-7365
x-mozilla-html:FALSE
org:Jefferson Lab;Physics Department
version:2.1
email;internet:wolin@jlab.org
title:Physicist
note:A218 CEBAF Center
adr;quoted-printable:;;12000 Jefferson Ave MS12H=0D=0A;Newport News;VA;23606;US
x-mozilla-cpt:;21408
fn:Elliott Wolin
end:vcard