|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The LIGO Executive Committee Agenda for Monday November 5, 2001 will be:
(Meeting time: 10:30 am Pacific Time)
Open meeting 10:30 - 11:30
Special Items:
This week we achieved the
first full-power lock of the Thermal Noise
Interferometer.
From: Rainer Weiss <weiss@ligo.mit.edu>
to: LSC executive committee
from: R. Weiss October 31, 2001
concerning: Notes from the LSC Executive committee October 26, 2001
LSC NSF ITR Proposal Activity
------------------------------
Reported on a conversation with Marvin Goldberg at the NSF. Members
of the LSC
are strongly encouraged to apply for ITR funds. The recommendation
is to have
people currently not supported by the ITR program as PI and to propose
around a new idea such as real time gravitational wave detection in
correlation with other astronomies. The proposed work needs to include
computer scientists and the application of innovative computing techniques
and/or networks. There are three levels of program for which one can
propose
each with different proposal deadlines. It looks difficult to organize
a major
proposal from LSC members in time, however, several smaller proposals
due
in the beginning of February 2002 may be generated by LSC members.
Assay of Phd Students working on LIGO I
----------------------------------------
Sam Finn and I have surveyed the LSC for on going PhD thesis research
on LIGO I
being undertaken by students associated with the LSC research groups.
The student's
names,institutions and project abstracts will be posted on the LSC
web page. There are
four experimental and eleven theoretical/data analysis thesis projects
currently
on going in the collaboration. The count only includes work directly
related to LIGO
I.
E7 run
------
The LSC Executive committee approved the request by Benoit Mour to
take part in
the E7 run. Benoit was encouraged to seek endorsement by the VIRGO
project
for his participation.
Report from the Software Coordinator
---------------------------------------
The final draft of the LSC White Paper on Data Analysis is circulating
through
the writing committee and will be sent to the LSC Executive Committee
in the next
two weeks.
An issue that needs further thought by the Collaboration, brought to
a focus
by the White Paper, is the way to deal organizationally with the new
initiatives
such as ITR and GRID computing. Currently the LSC is dealing with these
in an ad hoc
manner.
MOU proposed by VIRGO on collaborative data analysis by GEO,LIGO and
VIRGO
--------------------------------------------------------------------------
The Committee encouraged the Laboratory to go the next step with VIRGO
in
formulating the MOU. The MOU needs to include the role of the LSC in
the
data analysis
Overture to University of Wisconsin @ Milwaukee by individuals in the
VIRGO project
-----------------------------------------------------------------------------------
Bruce Allen told the committee of an overture that had been made to
carry out
joint data analysis by some members of the VIRGO project with UWM.
Although there
is no intent to regulate scientific exchange between individuals, the
committee
felt that an arrangement of such importance to the field and to both
LIGO and VIRGO
should be made formally between the Laboratories.
Publication on the planned advanced LIGO detector
-------------------------------------------------
David Shoemaker raised the issue that there is not now a publication
describing the
planned advanced LIGO detector which can be used as a reference for
the field.
The discussion that followed was in favor of the concept but did not
converge to
a plan. The issue will be raised again at the next meeting.
Next LSC Executive Committee meeting
-------------------------------------
Friday Nov 16, 2001 at 11:00AM ET
WBS 1.2 LIGO Operations--Administration
There was a teleconference on Thursday, November 1. Items discussed included:
From: Ed Chargois <chargois_e@ligo.caltech.edu>
>From: Linda Turner - turner@ligo.caltech.edu>
Web pages for the DCC give simple how-to's for document numbering, easy access to the latest on-line documents, and search capabilities for the DCC database. Take a look. . .
ACCOMPLISHMENTS
ACTIVITY
| Packages | Faxes | |
| In | 45 | 31 |
| Out | 9 | 35 |
Press here to access the DOCUMENT CONTROL CENTER WEB PAGE.
From: Esther Cunningham <esther@ligo.caltech.edu>
Press here for ACCOUNTS PAYABLE HISTORY DATA.
From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>
From: irena@ligo.caltech.edu (Irena Petrac)
SUPPORT (Wood)
Progress Period from 10.26 to 11.01
Accomplishments:
Weekly Advanced LIGO Project Controls meeting not held this week. Per Gary, we are scheduling our next meeting for November 13th. Continue the process of preparing material that will allow management to review the current status of the MRE effort.
Advanced LIGO MRE Proposal (Highest Priority)
Continue to update the TNI Schedule and incorporate any changes.Continuing the development of the Cost Estimate. Continue to pursue all other sub system schedules. Updated web site with latest data. The following is a summary of status by sub system: COC - Work on action items per second review continues. Next meeting date has not been scheduled. Attended meeting with LASTI and SUS to discuss schedule issues. PSL - Issued cost and resource loaded reports and awaiting comments. Next meeting date has not been scheduled. AOS - Work on action items continues. Next meeting date has not been scheduled. IO - Work on action items continues. Next meeting date has not been scheduled. SUS - Working with SUS group to incorporate current comments. Next meeting date has not been scheduled. Attended meeting with LASTI and COC to discuss schedule issues. SEI - Work on action items continues. Next meeting date has not been scheduled. Made schedule change as requested by David per his meeting with the SEI group. All other subsystems have yet to be scheduled for their first review. These dates are pending per direction from Gary Sanders and Dennis Coyne.
Project Web Site for posting schedule and progress related data continues to be updated with the latest and greatest.
Schedule 11.02 to 11.08:Next weekly Advanced LIGO Project Controls meeting will be November 13th at 1pm at the SSCR. Will continue to prepare and follow up with system leaders.
Advanced LIGO MRE Proposal (Highest Priority Task)
Will continue the development of the Advanced LIGO Project Controls Guide Book.
WBS 1.4.1.2 Project Controls (LIGO Construction)
Nothing to report.
The following change request has been distributed to the
Change Board:
| CR-010011
Revision A |
Construction | Adjust LIGO Construction Budgets to Reflect Actual Costs for Completed Tasks as well as REU Expenses, Accounting Adjustments, and 7-LIGO (old work order) costs (net $67K from Construction Contingency) | P. Lindquist |
Press for the latest Contingency Needs Projection.
From: Cindy Akutagawa <cindy@ligo.caltech.edu>
From: Ed Jasnow <jasnow@ligo.caltech.edu>
No report this week.
General Items:
--------------
(F. Raab)
Commissioning on 2 interferometers is the main concentration. The laser-safety
interlock installation by Moon Security is nearing completion.
2k IOO Periscope installation
-----------------------------
(C. Gray, B. Weaver, S. Whitcomb, and R. Savage)
We installed the new periscope on the 2k IOO/PSL table yesterday.
The
installation went very smoothly thanks to all the hard work and
attention to detail by K. Mailand, M. Smith, et al. The new periscope
performs about as expected. A before-and-after plot of the frequency
noise measured by the modecleaner (MC_F) can be found at
http://apex.ligo-wa.caltech.edu/~rick/Temp/NewPeriscope.pdf The
reference traces were recorded before swapping the periscopes.
The bottom line is that the peak that was at about 100 Hz is now at
about 22O Hz, in reasonable agreement with modeling. The narrow
features in the broad periscope peak appear to be narrow acoustic
sources coupling more efficiently due to the enhanced susceptibility
near the periscope peak.
We have received the second periscope and will begin making plans to
install it on the 4k IOO/PSL table.
No report
| Installation&
Commissioning:
Livingston |
Other Science/EngineeringActivities:
Issues/Concerns |
See also the Installation web page
The new periscope was installed and the beam repointed into the vacuum system. The results are similar to what had been observed at LLO; namely, the 100 Hz resonance in the MC_F spectrum has been moved to about 230 Hz, where the MC servo has greater suppression for frequency noise from the PSL.
Several more runs have been taken with the x arm locked for long durations to get data dor calibrating and tuning the tidal compensation.
The full interferometer cannot yet be locked during the day. The power recycled Michelson and a single arm cavity (state 3) can be locked during the day but not robustly. The interferometer may require active angular alignment control to stay in lock during the day and additional wavefront sensor servos are being prepared.
The common mode servo has been installed and is being tested.
Late news: Last night the feed forward system to compensate for the microseismic noise was tested on the recombined interferometer. The system worked well reducing the micrsoseism by at least a factor of 10.
Nergis elaborates:
Most of the effort this week has been devoted
to improving the power-recycled interferometer locking performance. The
two significant area of progress were:
(i) Increasing the frequency of successful lock
attempts: now the interferometer typically acquires within 1 to 2 minutes.
To do this we fine tuned RF phases while locked in State 5. We also remeasured
g-ratios in State 3, and this time the measured ratios were very close
to the ones we had been using to acquire lock, which had been arrived at
with a mixture of direct measurement and informed (we like to think) adjustment.
(ii) Extending the lock durations, which are
now typically 10 to 20 minutes. Here, closing feedback loops to the ETM
differential pitch and yaw from wavefront sensor 1 (at the antisymmetric
port) helped the most. Being able to better adjust the RF phases in State
5 to remove large spikes from the arm common-mode/laser frequency in the
PRM loops has also helped in keeping the sideband power in the PRM above
the servo shutoff level.
Other interesting phenomena: we have also been able to measure the mirror displacement due to the radiation pressure when the carrier power builds up in the interferometer.
Other persistent mysteries: we still see peculiar angle-dependent reduction in the sideband power build up. The present favorite theory is the occurence of an accidental overlap of the rf sideband with a higher-order transverse mode of the arm cavity, but we have yet to measure it. We also have some evidence of 60 cycle noise that seems to vary with time.
Mike Zucker
participated in commissioning on the L4k. Besides gratifying progress
on locking the full interferometer (reported elsewhere), did a number of
qualitative and quantitative diagnostics on RFI and 60 Hz ground currents.
This phase shift is added to the delay due to the Schnupp asymmetry. Such delay is weighted by the sign of the modulation frequency so that the combination is different for the two sidebands.
Another effect is the impact on the geometry of the mode and it turns out that an important contribution of Legendre[2,0] is excited. For this second part of the problem I am missing a constant, that is my predictions are proportional but larger than the observed values. For the TEM00 mode my simple model and data are exactly the same.
The Schnupp asymmetry that makes the dark port maximum for the sideband is the one that minimizes the imbalance between the sidebands. If a microscopic offset that compensates for the distortions is added to the vacuum lengths between the beamsplitter and the mirrors there is a large reduction of the imbalance since this cancels the impact on the TEM00 mode but not the excitation of the Legendre[2,0]. This excitation also depends on the degeneracy of the cavity but in a way I am still working on to explain it by analytical expressions.
Rick Karwoski
Frequency Reference boards:
1) Broadband noise noticied on AS I&Q signals in LHO 2k ADC readings (read from first Pentek in LSC crate). Center frequency of noise moves with different phase settings and number of TP selected. This noise does not show up on ADC channels of second Pentek board. Indication is that noise center freq. moves with change in servo loop calculation time. At Sigg's suggestion, first moved the sin/cos calculations required for phase shifts from the lsc front end to the isc supervisor. The front end that just performs a matrix multiplication and compute time of servo is constant. This stopped the noise center freq. shift when the phase shift was adjusted, however it did not remove the noise and shift in noise freq. still noted with number of TP selected. Since the ADC is externally clocked, change in lsc loop calc time should not, in theory, cause ADC noise, unless the loop was running so long as to miss ADC readings, which was not the case. The cause was finally tracked down to be the result of polling the ADC data ready bit in the ADC module control register. The code polls this bit on the first ADC board to determine when it is time to read in the ADC and compute the next servo cycle. When the code was modified to poll the second ADC card instead of the first, the broadband noise went away on signals from the first ADC and appeared on signals from the second ADC. This indicates some sort of ADC hardware problem. Unfortunately, removing the polling and rather waiting for an interrupt from the ADC for data ready adds a 10usec interrupt service delay to the loop, which is too long. Pentek has been notified of the noise problem and we are looking into alternative means of determining that ADC data is ready.2) LHO2k ASC experienced comms problems with one mid station controller. Problem was tracked to a faulty reflected memory board on the ASC front end (local processor could not write to it and all data coming in was corrupted). Reboot of system did not clear problem, however a power down reset did.
- Added code to systems which use Pentium processors and reflected
memory boards to automatically detect modules and addresses on code startup.
We have at least four models of Pentium boards with up to four PCI buses,
on which the reflected memory modules reside, and each wants to map the
modules to different buses/addresses. I had been keeping up with this in
header files, but this was becoming too troublesome. New code detects which
PCI bus the RFM boards are on and then returns the address pointers. This
makes the code more portable.
- Rep from Pentek was out last Friday. Data from our testing was passed on to them, particularly the noise measurements. They are supposed to look into the noise problem noted above, why DAC channels 5&6 on all our boards have higher noise, and the data packing problem we noted over a year ago.
- Updated the LHO2k ASC software. OptLev servo code was removed from the ASC front end and a newer release of end station front end code was loaded.
Luca Matone
Rich Abbot
Mohana Mageswaran
I have finished my photon calibrator new design and had given out for
PCB layout. I am ordering the parts needed for the project. I have been
also looking into the VME based stuff in order to come up with a VME interface
board.
Szabi Marka
I have a much more advanced version of GPS time monitor working and
I released a new version of the IRIG-B decoder/monitor code (DMT).
The following upgrades have been implemented:
1. The frame size has been increased from 1 sec to 8 sec or even 16
sec. This drastically reduced the overhead involved in opening and closing
files, as well as reducing the inefficiency involved in accessing small
blocks of data in a large number of files.
2. (Almost) all CDS Unix machines have received a memory upgrade. The
machines in the control room now have between 786MB and 1GB of RAM. This
allows to run data viewer and dtt in parallel as well as having a large
number of medm/netscape/terminal screens open at the same time.
3. The networking of these machines has been changed from 10bT to 100bT
by adding fast ethernet ATM equipment to the 3810 switches.
4. Part of the ldas high performance disk system (T3) has been directly
attached to a SUN Blade 1000 through fiber channel. This machine mounts
the ldas disks using the QFS file system and is currently being used as
a test frame builder.
5. DTT and data viewer have been updated to take advantage of the new
long data blocks. The dust hasn't quite settled yet. Stability still needs
some improvements but this should be ironed out in the next week or two.
This week we achieved
the first full-power lock of the Thermal Noise
Interferometer. All three cavities,
the Mode Cleaner and the North and
South Arm Cavities, now lock to carrier TEM00
modes and hold for hours at a
time, even during the day. Lock acquisition
is not prompt. It takes a few
hours to get both arms locked in TEM00 modes,
but once locked the system is
quite robust.
The noise in the North Cavity is, as anticipated, not substantially
different from the old configuration, remaining in the low 1e-17 m/rHz
range around 1 kHz, but the South Cavity is somewhat quieter.
The
equivalent length noise in the South Cavity between 700 Hz and 1 kHz
is
consistently 9e-18 m/rHz during the day, and one measurement at 10:00
pm
gave 4.7e-18 m/rHz at 1 kHz. Cross couplings between the cavities,
a
problem that had plagued the TNI under the old configuration, have
been
eliminated at the level of our preliminary measurements. We have
no noise
data from the South Cavity under the old configuration to compare this
with, since we were not able to reliably lock it before.
We are currently working on improving our lock acquisition time, and
after
that is reduced we will try to find out why one cavity is quieter than
the
other.
LASTI (Bayer, Fritschel, Goda, Harry, Laliberty, MacInnis, Mason,
Miller, Mittleman, Ottaway, Phinney, Rollins, Shoemaker, Zucker)
=======================================================================
BSC SEI INSTALLATION:
Myron, Fred and Bob have finished mods to the cartridge cleanroom to
customize it for installing the LIGO I stack. The BSC side door
has
been removed and the BSC dome was balanced & rigged for removal
today.
We have a few last minute odds & ends but we should be ready for
next
week's visit by LLO experts Gary Traylor and Harry Overmeier.
They
will help us drop in and assemble our downtube and stack elements.
HYDRAULIC ACTUATORS:
Ken is continuing to iterate the counter-wound spring design with
feedback from the Stanford group.
Joshua got his pump running and circulating oil, and is performing
tests to characterize it.
With some added shaker equipment shipped up from LLO by Joe Giaime,
Rich, Myron, Bob and Joshua repeated the BSC pier resonance test and
this
time achieved a clean result. First relevant mode turned up around
40
Hz, which is good from the standpoint of providing a robust reaction
impedance for the hydraulic actuator.
CDS/DAQ/GDS:
No progress on DAQ, GDS or RAID problems (MZ out of town)
PSL:
Dave, Jamie and Go have measured the actuator transfer functions for
the EO phase corrector and the PZT and they look pretty good (alas,
we're having trouble getting them from our ancient network analyzer
into Matlab for quantitative fitting...)
Simulation and Modeling (Bhawal)
E2E Physics Meeting
-------------------
Biplab explained how he's studying the mode mismatch problem while
the interferometer goes
from cold state to hot state. Then Erika discussed her FFT study of
the amplitude
mismatch problem of sidebands possibly due to misalignments in the
recycling cavity.
Lock acquisition & thermal lensing
----------------------------------
Biplab has started studying the effects of mode-mismatch as the interferometer
goes from cold
to hot state by mimicking the effects of thermal lensing by an effective
change in refractive
index of input mirrors. The runs suggest that it's necessary to change
the gain parameters
several times during this process to keep the system properly locked.
With help from Matt, he is trying to properly lock each temperature
state systematically.
CVS
----
Ed Maros added anonymous cvs read-only accesses for e2e software
Alfi
----
(Bruce)
- Completed edit capabilities of connection widgets.
- Work on updating of the container tree view.
(Melody)
- Continued on member node widgets.
- Worked on adding primitives onto the canvas.
LIGO Data Analysis System
Software Systems (Blackburn)
The number of bugs in the released version of LDAS was larger than
we
expected. This was especially true of basic functionality in the frameAPI
and the metaDataAPI. Steps have been taken to insure more regular routine
testing of all LDAS functionality. We will be setting up scripts which
exercise
all testbeds once every other night on the LDAS development system
in order to ensure that
problems can be detected earlier than they have been in the past.
The ILWD library was enhanced to support a new data attribute called
meta-
data. It is primarily intended as a place holder for information used
by
the LDAS system.
The frameAPI, dataConditionAPI, lightWeightAPI are all being extended
to
support missing frames (data gaps) in a user request for data. This
will involve a
slightly more structured exchange object for time series data and support
for missing data fills in the data ingestion and algorithm action sections
of the dataConditionAPI.
The new library (IlwdFCS) used to format data into FrProc frame files
that
was used by the dataConditionAPI in the 0.0.21 release is being ported
to
the eventMonitorAPI.
The dataConditionAPI, eventMonitorAPI and frameAPI are being extended
to
support a new set of user options allowing user defined data products
to
be written to a single frame or to separate frame files.
A proposal for a frame file naming convention has been reviewed and
circulated. A final draft is nearing completion. The conventions specified
in this document have been adopted by the development versions of both
LDAS and the framebuilder.
A quick assessment of the new GCC 3.0.2 compiler has determined that
it is not
ready for use in LDAS.
The enormous size of the disk cache at Hanford introduced a major process
inefficiency into the frameAPI. This has been sorted out and the improvement
in efficiency will become
available in the next release of LDAS well in time for the engineering
run.
Supported development efforts on GriPhyN underway at UWM for the upcoming
SC2001 conference in Denver the week of 11 November.
A new feature was added to the managerAPI that generates a timing profile.
This enables a
user to trace the performance of his jobs through the system. Also,
the wrapperAPI was
modified to provide additional timing information which is written
to the log files.
We assisted the Periodic Search Group in tracking down a problem in
their
code which was preventing them from being able to run preliminary search
codes in the LDAS system.
We discovered that the new LAM MPI library uses the SEGV signal. This
signal
is actually being thrown in the LAL/LALwrapper code to identify memory
management issues thereby introducing a risk of confusing the source
of
a serious signal in the OS and the signal handler used to process the
signal. The issues have been reported to the MPI working group. One
of our
Caltech software engineers has been tasked with looking into the LAM
signal handler's detailed behavior.
Hardware Systems (Anderson)
Caltech
-------
(Dan Kozak)
Helped move LIGO T3s from Synchrotron to Booth
Expanded SFS volumes in HPSS that had filled up
Worked on HPSS/hsi upgrade:
compiled hsi 2.6.2 for Sun,
HP V-class and SGI architectures
and installed it on lugh,
skinner/mulder/scully, earnest &
johnny.
assisted with HPSS software
upgrades and metadata conversions
labeled and imported more
9940 tapes
configured Ultra 10 (acsls.cacr.caltech.edu,
used for silo
management) per CACR network
and security guidelines, brought it
up on the network
Worked to get StorageTek to agree to repair a broken part on Powderhorn
silo under warrantee.
We are still awaiting the repair to be performed...
(Al Wilson)
Getting ready for RH7.2 have preped a directory for the RPM's Checked
out
and have a final set of rpms for RH7.1 - 2.4.9-6 kernel. The ldas-test
system will be the first installation to be up graded to the new kernel.
also tested a test start config that will not delete the usr1/lcldsk
directory.
(Stuart Anderson)
Continued testing of RedHat 7.1 updates and RedHat 7.2 for use as next
LDAS
Linux reference platform.
Contributed to the new LIGO Frame file name guidelines.
Helped with configuration and testing of QFS file system at LHO for
the
CDS framebuilder.
Received in-situ fiber length measurements for existing LDAS single-mode
fiber runs at Caltech. The lengths are short enough that we will
install 50micron multi-mode fiber side-by-side to allow the use
of cheaper computer equipment when available. If this configuration
works in practice at Caltech we will specify 50micron multi-mode fiber
for the network connections to the new buildings at LHO and LLO.
Moved the LIGO<->CACR ATM OC-12 connection from a test workstation
to the LDAS-DEV gateway box for increased bandwidth in archiving
and accessing Engineering run data.
Reinstalled and configured Solaris on the form GC computer "sargas"
to
prepare the system for use as a test machine to run and test the initial
LIGO archive system until the new Sun servers are delivered.
Livingston
----------
(Shannon Roddy)
Admin locked up yesterday evening and did not respond at all to
keyboard input, pings, etc. I have not had too much time to
investigate, however Stuart has recommended that I install the newest
set of patches.
BigBrother & Tripwire: I found a third party script that has
been written for BigBrother to interrogate tripwire. I am checking
into the feasibility of using this rather than writing my own
using the examples in SysAdmin magazine. Fixed some problems
with
bigbrother and rsync - Thanks Stuart. Will have to reinstall
on ldas-sw
to get the cgi working again.
Ordered and received the QFS licenses for LDAS and CDS. Still
working
an issue with the backup scripts & secure shell.
Hanford
-------
(Greg Mendell)
1) QFS was installed and configured on the LDAS dataserver and the CDS
fb3 framebuilder. Initially, kernel messages indicated a problem
occurred on fb3, e.g.,
"fb3 samfs: [ID 145447 kern.notice] NOTICE: SAM-FS: sam_wait_space:
/fb3qfs1: File system full - ENOSPC".
Removing the mount option "forcedirectio" seems to have solved this
problem, but testing continues.
2) Calculated tapes needed for E6 run, and started plans to archive
data
for this run.
3) Debugged the knownpulsardemod DSO so that it now successfully
produces SFT frame output when run as an LDAS job. Started planning
the
known pulsar MDC.
General Computing (Wallace)
MIT:
-Upgrading admin assistant machine from win98 -> win2K
-Moving accts over to new file server (piece meal)
-Adding accts to modem pool
-Investigating licenses for win2k and acrobat
Livingston:
(Shannon)
-Fixed a couple networking issues. Moved a user over to IMAP
instead of POP. This will help in keeping his mail synced between
home
and LLO. Swapped out a laptop for a user to replace his older
laptop.
Reenabled a user account on the GC network.
Hanford:
- As of Wed. 10/31/01, we are running on the new NIS+ server,
still
called rainier but it is now a new Sun E220 instead of the old E3000.
This has put more emphasis on our application server as all application
software has moved off the NIS+ server. The swap went quite well.
The
users are still identifying application software, mostly freeware,
that
used to exist on the NIS+ server and now needs to be installed on the
application server. Before anyone can change their password on
the new
NIS+ server they must "reinitialize" their password with the server.
The steps for this process have been published.
-Setting up a new PC to handle the video-conferencing camera.
CIT:
(Mike)
-Hardware problems with cdrom & zip drive repaired on a PC. I have
corrected
this problem and user is back up and running with full functionality
of the
computer.
-I performed a complete rebuild with OS and all General Computing software.
-Swapped out the laptop, for Ed C., to a more updated and lighter Laptop
this
included moving all his personal files over and setting network configurations.
-Ghost Enterprise Network is up and running, the biggest scenario to
this
software is configuring the different network cards on the project
getting
the correct dos drivers for each nic card to work; so far it looks
like
there is four different network cards for the PC's; now I am starting
to
look into the Laptops. This is going to be a great tool for the General
Computing a big plus for the project. If a users PC gets corrupted
or hard
disk failure I should be able to recover within 30 minutes or less
depending on the size of the image. This will be done without moving
the PC
to a staging area, this process can now be implemented threw the network.
We will also be configuring this software to push additional software
packages and security patches threw the network but not to bog down
the
network causing a bottleneck.
-Reloaded 2 PC's with OS and all General Computing Software.
-Loaded Rack Mounted server with NTSRV 4.0 and am going to run server
base
Mcafee software, plus additional security packages.
(Bruce)
Gnats maintenance:
(1.0 days)
- Completed set up gnats
for the detector hardware group.
(Lisa)
- Began looking at wireless security.
- Worked on some secure shell tunneling problems.
- Resolved a problem with solidworks on sunpci. It looks like
running NT
instead of win98 will fix the display issues.
- Rebooted sirius. We have had a chronic problem with ultra10's
locking up
after sirius gets a reboot. It looks like MU5 resolves many of
these problems.
- Built a development apache/mysql/php server. This was done
using the solaris
8 bundled apache/perl. I have started working on putting together a
web based
system to manage our IP & inventory information. This is
becoming a
necessary tool. There is no easy way right now to manage things
like mac
addresses or tracking computer moves.
-Onsite user and phone support that included software, OS problems
and
printing problems.
(Larry)
-Worked on updating a number of programs on the servers. Moved a number
of
the servers around in the server room in preparation to swap equipment
out of
the racks.
-Contacted NoaNet concerning connectivity for the Hanford Observatory.
They are
working on getting more information sent to me. When I receive it I
will pass it
along.
-Resolved a couple of e-mail issues. The tools being used by everyone
all have
different methods of configuration which have caused some conflicts
for the user
but most issues have been easily resolved.
-Working on the virus-scan s/w for the e-mail server. The first tests
have not
proven too positive for the virus-scan but the smtp check is looking
pretty
good.
-Worked a number of procurement issues. Mostly, trying to get things
corrected
with SUN.
-Working on the fiber installation for the GigE backbone. The distance
for each
location from Bridge Annex has been measured; WH 2400 ft., 40M 2300
ft.,
Millikan 613 ft., Lauritsen 1448 ft., Booth 1320 ft. .
-Spent a little time working on security policy documentation.
-Some time working on training and getting more of the WEB duties moved
to
others. Setup a couple of new WEB procedures for different people.
From Calum Torrie:
Advanced LIGO
Modelling and Testing for proposed Prototype Suspensions
Continuing to add to and adapt existing engineering drawings for prototype
mode cleaner suspension.
Started to design the upper mass with particular attention to mass
and moment of Inertia and position of centre of mass
I have been looking at wire suppliers in the USA. Planning on getting
some spring steel wire sent to me from California Fine Wire Company to
test.
Started discussions this week with Mike Plissi and Eoin Elliffe in GEO
and Janeen here at Caltech to model the cantilever blades we will use on
the prototypes on I-DEAS. In particular we are interested in the deflection
and internal mode of the blades.
I hope to build up a small test facility in order to test some prototype
blades here in the lab, and also as I mentioned before to test some wire.
From Mark Barton:
This week I finished incorporating the thermal noise according to the parameters in Bench plus corrections from Peter Fritschel and Phil Willems. I also made some optimizations that halved the run time for the short version of the calculation (without thermal noise) that will be used for tweaking the normal mode shapes and frequencies, and reorganized the code to make it possible to try different damping models without having to redo the time-consuming part of the calculation (finding the second derivatives of the wire bending potential terms).
For additional information about this report, contact sanders@ligo.caltech.edu