|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The LIGO Executive Committee
Agenda for Monday March 8, 2004 will be:
Special Items:
no report
Status of LSC/MOU Research Updates
and Program Reports (Petrac)
ACIGA:
Attach. B for 08-15-03 to 02-15-04 / signed-off
Attach. C for 08-15-03 to 02-15-04 / signed-off
Attach. D for 08-15-03 to 02-15-04 / signed-off
Attach. Z for 08-15-03 to 02-15-04 / signed-off
AURIGA:
Draft-MOU and Attach. No. 1 / in LIGO review / Barish and Sanders
Carleton:
Attach. A for 02-15-04 to 08-15-04 / in LIGO review / Barish
Attach. Z for 02-15-04 to 08-15-04 / in LIGO review / Barish
Dominguez Hills:
Attach. A for 02-15-04 to 08-15-04 / in LIGO review / Lazzarini, Weinstein,
Riles
Attach. A for 02-15-03 to 08-15-03 / in LIGO review / Barish
Attach. A for 08-15-03 to 02-15-04 / in LIGO review / Barish
Attach. Z for 02-15-04 to 08-15-04 / in sign-off by PI
Goddard:
Attach. A for 08-15-02 to 02-15-03 / in LIGO review / Barish
Attach. C for 08-15-02 to 02-15-03 / in LIGO review / Barish
Attach. D for 08-15-02 to 02-15-03 / in LIGO review / Barish
Attach. Z for 08-15-02 to 02-15-03 / in LIGO review / Barish
MSURG:
Attach. A for 08-15-03 to 08-15-04 / in LIGO review / Barish
Attach. B for 08-15-03 to 08-15-04 / in LIGO review / Barish
Attach. D for 08-15-03 to 08-15-04 / in LIGO review / Barish
Attach. Z for 08-15-03 to 08-15-04 / in LIGO review / Barish
GEO:
Attach. A for 08-15-02 to 02-15-03 / LIGO review / Barish
Attach. B for 08-15-03 to 02-15-04 / in sign-off by PI
Attach. C for 02-15-04 to 08-15-04 / in LIGO review / Barish
Attach. D for 02-15-04 to 08-15-04 / LIGO review / Barish
LaTech:
Attach. A for 02-15-04 to 08-15-04 / LIGO review / Barish
Attach. Z for 02-15-04 to 08-15-04 / in sign-off by PI
LSU:
Draft-MOU for Giaime's LLO appointment / in legal review / Pool &
Jasnow
Michigan:
Attach. A for 02-15-02 to 08-15-02 / in LIGO review / Barish
Attach. A for 08-15-02 to 02-15-03 / in LIGO review / Barish
Draft-Attach. A for 02-15-03 to 08-15-04 / in process for web posting
Attach. A for 02-15-03 to 08-15-04 / in LIGO review / Barish
Attach. Z for 02-15-04 to 08-15-04 / in LIGO review / Barish
Rochester:
Draft-Attach. A for 08-15-04 to 02-15-04 / in LIGO review / Lazzarini
SLU:
Attach. A for 02-15-04 to 08-15-04 / in LIGO review / Barish
Attach. Z for 02-15-04 to 08-15- 04 / in LIGO review / Barish
Southern University:
Attach. C for 02-15-04 to 08-15- 04 / in LIGO review / Barish
Attach. Z for 02-15-04 to 08-15- 04 / in LIGO review / Barish
Trinity:
Attach. A for 02-15-04 to 08-15-04 / in LIGO review / Barish
Attach. Z for 08-15-04 to 08-15-04 / in LIGO review / Barish
USC/ISI:
Draft-Attach.1 for LIGO data / in LIGO review / Barish
Washington State:
Attach. A for 08-15-03 to 02-15-04 / signed-off
Attach. A for 02-15-03 to 08-15-03 / signed-off
Attach. Z for 08-15-03 to 02-15-04 / signed-off
Attach. Z for 02-15-04 to 08-15-04 / signed-off
VIRGO:
Draft-MOU and attachments for joint work on data analysis / in VIRGO
review
University of Washington:
Proposed collaboration on managing angular instabilities in Fabry-Perot
cavities
Draft-MOU / in review by the PI (since Nov. 2003)
Draft-Attach. D / in review by the PI (since Nov. 2003)
Draft-Attach. Z / in review by the PI (since Nov. 2003)
Note: Per recent telecon UW's response to above is in the works
Wisconsin:
Attach. A for 02-15-02 to 02-15-03 / in LIGO review / Barish
Attach. A for 08-15-03 to 02-15-04 / in LIGO review / Barish
Attach. Z for 02-15-03 to 08-15-03 / in sign-off by PI
LIGO Weekly Site Telecon (Lindquist)
A site teleconference was held on Thursday, March 4, 2004. The
following were discussed:
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
No report this week (vacation). Will be at conference next week
and at the LSC Meeting the week of March 15, 2004.
>
From: Cleveland Mak <mak_c@ligo.caltech.edu>
ACTIVITY
| 03/04/04 | Packages | Faxes |
| In | 20 | 25 |
| Out | 7 | 26 |
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)
[pel] University
of Oregon: The contract for on-site support at LHO is under review.
We have received a final invoice. Current year costs are being moved
from the old Visitor's Grant to make room for the invoiced amount and the
invoice is being paid.
[pel] MIT:
A contract mod has been prepared and processed for payment of latest invoice.
From: Ed Jasnow <jasnow@ligo.caltech.edu>
SUPPORT (Baldon, Lloyd, Tischler)
>Irene Baldon
No report this week.
>Dorothy Lloyd
ADVANCED LIGO (Cost Schedule
Control Systems) T. Frey
From: Thomas Frey <tfrey@ligo.caltech.edu>
For list of documents that are being used to develop Adv. LIGO Cost and Schedule, see http://www.ligo.caltech.edu/~tfrey/Cost_MTG_082002/
Finished input of added cost sheets and changes to cost sheets.
Posted revised plan and progress schedule. This includes earned value analysis.
IO - Still need completed WBS dictionary and BOE.
Completed cost book changes per Jays comments.
Will post revised plan and progress schedule.
LDAS - No action required at this point in time.
INSTALL - No action items pending.
DAQ - No action required at this point in time.
PM - No action required at this point in time.
FAC - No action required at this point in time.
SUP - No action required at this point in time.
LASTI - Will input changes per teleconference with David Ottaway and Carol W. on the 20th.
Followed-up with Irena regarding reported challenges with LSC mail lists.
Continue to work with Carol and Florence starting work setting up "crosswalk" between accounting and progress reporting.
Project Web Site for posting schedule and progress related data continues to be updated with the latest and greatest.
Tom Lucas, the LIGO video producer, was at MIT this week where he met
with key LIGO personnel. Tom will be visiting Livingston during the
LSC meeting. We will try to show him some gee-whiz stuff and set
up some meetings with key LSC personnel.
Change Requests:
| CR-030015 | FY 2003 Livingston Observatory Detector Maintenance Expenses (Increment) | R. Wooley | July 14, 2003 |
| CR-030016 | Hanford Facilities 2.2--Divide the Large Equipment Access to Facilitate Movement of Large Items (currently assigned to FY 2004 liens list) | J. Worden | July 31, 2003 |
| CR-040002 | Added personnel--Hanford | P. Lindquist | February 13, 2004 |
There was a meeting of the LIGO Change Control Board on Monday, February
23, 2004. Minutes have been prepared and distributed (LIGO-M040064-00-P).CR-040001
replaces CR-030015 and CR-030015 is closed. CR-030016 has been placed on
the "watch" list and is closed. Change Request CR-040002 for added personnel
at Hanford was approved and has been incorporated into the FY 2004 budget
baseline.
From: Cindy Akutagawa <cindy@ligo.caltech.edu>
No report this week.
Summary of Commissioning Activities at LIGO Hanford Observatory (compiled by M. Landry)
Thermal compensation studies continued, occupying much of the time on
the 4k
this week. Initial studies with a state 4 full lock occured on
Friday.
At this early stage, the assessment
was that the TCS was doing a reasonable
job of correcting the phase front curvature error in the PRM.
Later studies incuded asymmetrical PRM heating (only ITMY), and here with ITMX.
TCS modelling
has shown that 1cm offsets in the position of the central heating spot
on
the test mass are tolerable. Furthermore, the factor of ~3 more
heating required (than
expected) to optimize the sideband power (noted last week) was shown
to be not due
to this type of offset.
Other TC studies included an annulus
mask test, another state 4 heating
curve,
and another asymmetrical
heating study.
Using the Spiricon frame grabber to estimate the position of IFO beam
on the BS,
an alignment
servo was formed, feeding back to MMT2.
Recall that dust falling through the beam on ISC tables produces glithes;
a
HEPA filtering system was added to ISCT4 last Friday, intended for
cleaning after
an entry onto the table. Studies are underway to see if the filter
is effective cleaning
agent.
Violin resonances (both ITMs and ETMs) appear to have RMS a factor of
four
higher
in S3 than in S2. It is not easy to see how this could be the case,
given
the differences in ITM and ETM drive configurations (radiation pressure?).
An effort was made to increase the power out of the H1 PSL, and confirm
PMC loss
measurements
(which followed on some measurements earlier in the week of PMC
losses and settings).
---------------------------------------------------------------------
L1 commissioning (summary by MZ)
--------------------------------------------------------------------
This week was a mad rush to complete interferometer diagnostics
(mostly related to high power operation, thermal lensing, and
understanding the sideband buildup)
before we sealed the X end station off for the HEPI installation.
[Note: this is just a safety precaution, nothing is vented].
Unfortunately we were plagued by intermittent laser trouble; in the
end,
this prevented us from getting much useful data for the week. The fault
presents itself as a thermal overload in the NPRO master oscillator,
although it is not clear if recent cooling loop changes, a not-so-recent
increase in pump diode current, or a random unrelated
failure are contributing causes. We will investigate
further and determine if swapping in the spare MOPA is the best
way to go.
The good news: we sealed the X end station
Thursday morning and
HEPI actuator installation is under way! (more
below).
Education and Public Outreach (B. Wooley)
---------------------------------------------------------------------
LLO visited by 82 Seniors from Parkview Baptist. Accompanied
by their
physics instructor they enjoyed hands-science, a tour, and painting.
Safety (R. Riesen)
---------------------------------------------------------------------
The self-locking door handles on the 2 interior doors for the machine
shop
have been replaced with non-locking handles.
Installed a UPS for the Laser Safety/ security system computer.
Completed the disposal of the GSA auction equipment that was at the
"in-town" warehouse.
I have "on-site" equipment ready for GSA removal.
Working on finishing the punch list for completion of safety items required
in the HPTF.
Found no safety concerns during my weekly site safety tour.
Continuing work on LLO's property audit paperwork for Ed Chargois.
Property tagged all existing HEPI equipment with the new "LLO Seismic"
stickers.
Updated and revised the Laser Safety/ Security log-in and operation
procedures.
Attempting to streamline the shipping and receiving procedures.
I have spent most of today overseeing safety procedures during the
installation of the HEPI system at the X-end station.
Scheduled CPR training for LLO personnel for the 20th and 21st of April
'04. 15 people have signed up so far.
HEPI fabrication and installation (J. Giaime, R. Abbott & MZ)
---------------------------------------------------------------------
After a status review Wednesday afternoon
gave the green light, all
the X end alignment data were baselined and
verified by the installation crew.
The X end gate valve was shut first thing
Thursday. By 3 PM the entire
FAS/CAS package was gone from the first pier,
and the template was set
for drilling the new holes in the pier top.
The new precision lift jig and
procedure worked as advertised. Film at 11.
In other news, after persistent bad luck (including
galvanically incompatible pressure sensors and a
"surprise" connector pinout change by the valve
vendor) the valve calibration stand
is finally in business. Katrina and Tom are busily calibrating the
first 8 valves needed for the X end (just in the nick of time).
Piping contractors continue to beat the schedule; Ken Mailand is
working hard to keep ahead of them with connection details
and fittings.
L1 commissioning (Frolov)
---------------------------------------------------------------------
I worked on testing the upgraded daq system. I diagnosed a
problem with the summing of the signals from the lsc as port photodiodes.
This problem was likely the reason for 10x higher than nominal
interferometer noise in the low noise state.
With Andri's help I installed and successfully ran Rana's ifo noise
model
on my computer for the analysis of the ifo characterization data that
were collected after S3 run.
L1 commissioning (Gretarsson)
---------------------------------------------------------------------
Began working to make Rana's noise model more
accessible and easier to modify. Once the initial stage is done,
I will
post it on a website with instructions.
LDAS (I. Yakushin)
---------------------------------------------------------------------
LDAS admin:
1) The power panel's unfortunate property to physically disconnect
power
to LDAS in response to short outages has been fixed by Rusyl and Excel:
it is now powered through LDAS UPS.
2) Requested IP addresses for grid computers.
Data analysis:
1) Continue running simulations on the whole S2.
2) Preparing presentations for LSC meeting.
3) Trying to use Shourov's linear prediction filters with waveburst.
4) Regenerated all sky black hole merger MDC frames after Patrick Sutton
discovered that they were actually half sky.
5) On Erik's request working on generating MDC frames with events
injected not only at zero time lag but at time lags that are multiples
of 5 s.
General Computing (S. Roddy)
---------------------------------------------------------------------
General Computing (Shannon Roddy):
* Current bandwidth usage can be seen at
http://teche.ligo-la.caltech.edu/mrtg/LLO-Router/130.39.245.1_1.html
* Archived bandwidth usage can be seen at
http://teche.ligo-la.caltech.edu/mrtg/archive/ under the dated folder
for the week of interest.
* Finished moving all of the home directories to the new file server.
I
began moving them manually and realized it was going to take too long,
so I wrote a script to move the directories and update the NIS+ tables.
The script took 2 1/2 days to complete so I would have been there
forever moving the directories by hand. All of the home directories
are
now on the RAID server. Still have a few things to finish
up, but for
the most part this is complete.
* Set up two laptops for Bonnie and Katrina to use for two tours this
week.
* Set up a replica NIS+ server to get ready to move more stuff off of
our old file server. Had some issues in the middle of this, since
the
UPS that baronne is plugged into had some problems during the
replication. It caused NIS+ to crash, then the machine was not
stable
and I had to reboot it. Commands like ps would hang and never
finish.
Luckily, the NIS+ tables were not corrupted. The last time this
happened it took us over 12 hours to get everything back. Once
I got
everything stable and running again with the replica server running,
I
decided to leave things alone for a while.
* Promoted the replica server to be the master server, removed baronne
as a server, and then created a new replica server for redundancy.
The
old NIS+ server is now a client. The mail server (abundance)
is now the
NIS+ server and touro is the replica.
* installed a tape drive on abundance to back up the mail spools and
the
system. Set up a script to run ufsdump nightly to the tape drive.
Later this will be set up in a different fashion, but this will give
us
backups for now.
* still working on compiling more apps and moving the old ones to the
new file server. Need to make sure that when I compile things
that they
do not link to libraries in /usr/local since not all machines will
have
the same packages.
* set up another ntop server so that I can monitor bandwidth usage.
I
will later be setting up snort and ACID on the inside of the firewall.
The new Cisco 3550 works great for this and is much more flexible than
the old PIX. I can now span ports and use them for IDS purposes.
I
could not do this with the PIX. However, the PIX will be useful
when I
internalize the PCs.
* LSU contacted me about the old T1 router. I think it is safe
to say
that we will not need it anymore. I have emailed them and told
them
that I can return it since it is in their inventory. I also contacted
Ric Simmons and asked him to turn off the T1 lines permanently now
since
the GigE has been stable for the most part.
* Put in a request for 4 DNS changes for LDAS.
They are adding 4 new
gateways into the LDAS network. In my opinion
this violates the
"spirit" of the LIGO IT security plan, but there
is probably not much
that can be done about it. I count the
number of gateways to internal
networks to be at least 12 now, though I am sure
there are a couple I am
not thinking of off the top of my head.
This also puts the number of
operating systems at 4. Many of the gateway
machines do not have a
clear administrator responsible for the machines.
The only layer of
defense is the firewall and for some reasons
this is not sufficient at
this point. I am concerned that this will
only get worse.
#MZ I'd like to be copied on reactions to this
issue. I get
the distinct impression LDAS, GC, DMT and
CDS admins need
improved coordination lab-wide, but would
love to be reassured
if everything's truly peachy.
* Dell has provided a fix for the problems that I was experiencing on
the new Gigabit switch. I have not implemented it yet, but it
involves
a filter being put on our Foundry core switch. Hopefully it will
fix
the problem since I am only able to use 2/3 of the ports on the new
switch. So far I am more pleased with Dell's support than I am
with
their hardware.
#MZ hey, you can't have everything...
S2 Burst Analysis (K. Yoshiki Franzen)
---------------------------------------------------------------------
1) Finished my WaveMon veto study of the four interesting S2 burst events.
No L1, H1 or H2 WaveMon triggers have been found which can be used
to veto
these events. Please find the complete result at
http://www.ligo-la.caltech.edu/~franzen/restricted/WBveto4.html
LSC Burst group password required.
2) Finished processing the remaining S2 playground H1 and H2 data with
WaveMon.
We now have the complete set of WaveMon triggers from all three instruments.
I
am now proceeding by calculating dead time fractions and WaveBurst
veto
efficiencies.
CDS support (Ash Khan)
---------------------------------------------------------------------
1) Continuing the support and checking of the new DAQ system and upgrading
the procedures to reboot the controllers as we find new bugs.
2) Developing the software for the Laser power supply and the new Chiller
communication.
no report
no report
LASTI WEEKLY (Allen, Mason, McInnes, Mittleman, Ottaway, Sarin, Ruet)
HEPI
We have finished reinstalling the actuators onto the BSC. After releasing
all of the stops we found that the springs were not acting as if they
were
free (see LASTI ilog entry ilog for 2/27/04 entitled BSC - MOVING THE
PLATFORM WITH THE SPRINGS). It is likely that the springs are binding
in
the housings near where they are secured. We are waiting on new equipment
(to
replace what has been returned to Livingston) so that we can lift the
platform, release the tension in the springs and make adjustments.
This
should not be a problem at the observatory since the design
has been changed. We have some improved isolation results for the HAM
chamber, which are
posted on the LASTI ilog, 2/27/04 - HAM Performance Results. The main
change in the controls was in the STS sensor correction filter. This
filter is still limited at low frequencies by the tilting of the floor,
we hope
that there is less tilting at the observatory which will allow more
low
frequency gain and improved performance.
Electronics design and fabrication
During the month of February, the following tasks were completed:
1] the fab, assembly, test and shipment of phase camera chassis #2
to Hanford.
2] the fab, assembly, test and shipment of position sensor tester #2
to
Livingston.
3] the transfer of my office essentials to the lab to eliminate inefficiences.
4] the compilation of critical parts shortages in the lab for restocking
order.
HEPI VME Controls development
Developing and testing software for performing system identification
on
the HEPI-HAM using the VME controls system. The motivation for doing
this
is to test the infrastructure here at LASTI before we do system-ID
at the
LLO HEPI installation starting about March 15. The software is DTT
based,
i.e. uses DTT libraries for doing things like swept-sine measurements
and
talking to EPICS. The notable additions are that saturations of
measured channels are monitored, so that drives to actuators are reduced
whenever saturations are detected. In addition, a number of channels
are
tracked as 'safety channels' - a deviation from nominal of values of
any
of the safety channels trips a safety trigger and shuts down the actuator
drive.
Mode Cleaner Fit Check Structure
Janeen Romie has sent us a structure that will allow us to check the
fit
of MC Tripple suspensions to our Xend-Ham.
HEPI Installation
Ken spent two weeks at Livingston assisting in the preparation for
HEPI
installation. Myron is now there for two weeks assisting in the
installation at the x-end.
A finite element model was created of a new BSC lift plate allowing
the
stack to be lifted over the HEPI springs
Thermal Compensation Design
A meeting was held with the team which installed the thermal compensation
at Hanford. It was discovered that the mounts for the camera housing
was
inadvertently installed upside down resulting in problems with the
installation. A list of improvements was gathered for the camera
housing
and laser enclosure for future systems
Simulation and Modeling (Bhawal)
SimLIGO noise hunting:
-------------------------
(Matt) Mysterious noise source (in the model)
which dominates around 100 Hz found
to be the digitized REFL_I feeding into
the MICH and PRC loops
through off-diagonal elements. Because
of the CM loop closed on
it, REFL_I only explores one or two bits
and is thus a non-linear
noise source. Removal of the off-diagonal
elements (which did
nothing good) removed this noise source
and put the SimLIGO
seismic-only noise curve well below the
SRD.
Effect of Thermal heating on WFS 3 and 4
-------------------------------------------
(Biplab) The contribution of common mode misalignments from Recycling
mirror (RM), End mirrors (CEM) and Input mirrors (CIM) in the
reflected
port (non-resonant sideband frequency demodulated) Wave Front
Sensors
(No. 3 and 4) were calculated as a function of (i) Gouy phase
and
(ii) for different heated states.
Results show that proper isolation of degrees of freedom at either
WFS 3 or 4 needs a careful choice of Guoy phase in telescopes.
With heating and proper choice of guoy phases, RM signal can
increase
by a factor of 3. CEM signal does not change at all by either
common
or differential heating simply because sidebands do not get affected
by CEM mode of misalignment.
Sent all results to Luca and Daniel. Also results can be looked
at enote
[elog-like but subject-based not date-based; Recently introduced
by Bruce Sears]: http://ldas-sw.ligo.caltech.edu/enote
Code for Dual-recycled Michelson
-----------------------------------
(Hiro) The fast but approximate field evolution calculation in a dual
recycling Michelson cavity has been refined, and a code has been
written in matlab for quick prototyping and debugging. The debugging
is in the last stage. By the middle of next week, a preliminary
document and completed matlab code will be ready for release.
Code development and maintenance
----------------------------------
(Hiro) Modeler code revision:
Matt and Hiro worked to revise a template matrix class to implement
the expression template technique. Now, most of the matrix and
vector
calculations can be coded as is, like
E_front_reflected = FrontReflectionMatrix * E_front_in +
TransmissionMatrix * E_back_in
without worrying about the C++ operator overloading overhead.
The sparse matrix handling has been revised as well, which is
important for the state space model simulation used for the many
of
the mechanical subsystems.
Hayakawa of ICRR (Institute for Cosmic Ray Research, Japan) has
64bit machine in ICRR and has been serving as a
very good guiena pig. For the release of e2e-1.9.0, he found
a
problem related to 64 bit architecture, and we could fix the
problem
very quickly with his concise debug information.
(Matt adds) Matrix Expression Templates: new implementation allows
for
evaluation of any expression and can be specialized to evaluate
specific expressions optimally. This seems to be ideal.
(Melody) Modeler:
Working on some enhancements to the FUNC_X code.
Worked on the FUNC_X technical document to describe the use of
dynamic linking and shared libraries in E2E.
(Ed Maros) Worked on modifying installation according to Matt's
experiences.
Alfi
-----------
(Bruce)
- Redesign and implementation of connection building
object in preparation for completing bundler rework.
(Melody) Working on executing the user's editor (specifically vi
or emacs) as the default editor for the primitive settings.
LIGO Data Analysis System
Software Systems (Blackburn)
We've spent the last week and a half migrating to newer third party
software
development tools. These have been tested on a tandem system and are
being
moved onto the LDAS-DEV system today. This is primarily a maintenance
related
activity with only minor bug fixes being addressed by the newer packages.
The
list of upgraded packages is:
* cURL - 7.11.0
* coreutils
- 5.2.0
* binutils -
2.14
* gcc - 3.3.3
* lam - 7.0.4
* xerces - 2.4.0
* BWidget -
1.7.0
* tth - 3.33
* tkcvs - 7.2
We have begun the porting of LDAS and its dependencies over to the Solaris
C/
C++ compiler. So far only the third party software in LDCG required
to build
and run LDAS has been ported to the Solaris compilers. No issues have
been
discovered at this phase of the port.
We have finally "cracked the nut" with regard to upgrading to the new
DB2
database server. This is after over a year of frustration from lack
of DB2
support on RH 9. IBM is now supporting DB2 under Redhat Enterprise
and as
such has had to post several fixpacs and specialized install procedures.
We
took these RH Enterprise procedures and applied them to RH 9 and with
a bit
of juggling were able to find the combination of sequences to get the
DB2
version 8.1 to run with LDAS 1.0.0. This database is expected to provide
new table support that potentially can greatly improved query times.
It is
also expected to be about 20% faster all around. However, the real
pay off
with this version is the ability to Federate the LDAS databases! This
has
been on hold for over a year because the Java Administration tools
didn't
work properly under RedHat 9 prior to this week's effort.
Several diskcacheAPI related bugs have been discovered in the LDAS 1.0.0
release when migrated to the sites. These are being worked on and should
be resolved in a week or two.
Other APIs that have had bug fixes applied are:
* cntlmonAPI - fixed parsing of non-standard dso
* ligolwAPI - added code to handle ligolw time out and testing on tandem-iii
* eventmonAPI: - removed extra killJob call; added capture of thread
states
for dataRecv to debug missing objects from wrapper.
* cmonClient: - working on all pages to be created when as needed rather
than
at startp and to reset only as needed.
There are currently 175 non-closed problem reports on LDAS of which
166 are
software and 19 are hardware issues.
Hardware Systems (Anderson)
No report this week
(AL note) However, the status of the CIT production system can now
be viewed at the following URL:
http://ldas-gridmon.ligo.caltech.edu/ganglia/?r=day&c=CIT&h=&sh=0
This tool was adopted from the GriPhyN development effort and is being
used by a
number of grid projects with large clusters.
Data Analysis Activities (Lazzarini)
Mendell:
The StackSlide code has been updated to better estimate the noise to
improve calculation of SNR. (Options were added to remove peaks from
noise estimation; an option to use the running median will be added
once
the LAL code for this is ready.) I also added the option to output
a
Hough number count instead of StackSlide power. Validation and
benchmarking are going well so far, good enough to get preliminary
results for the LSC meeting. Much more work will be needed (along with
Monte Carlo studies) to get final results. I also added updated
documentation in README_ComputeStackSlideSums and
README_CommandLineArguments to the lscdocs/pulgroup/stackslide cvs
archive; eventually all will get moved to LAL and LALapps.
Shawhan:
* Prepared report about Matlab for the Data Analysis Software Working
Group (DASWG).
* Generated list of S3 time intervals, for each interferometer, during
which
pulsar injections were on AND the interferometer was in science mode.
* Tracked down a recent problem with conlog as being due to the fact
that
the /cvs filesystem on the LHO CDS cluster filled up, causing file
corruption. Repaired the files and index to a self-consistent
state.
Sylvestre:
o Worked on running tfclusters on the full S2 data.
o Checked that the Poisson hypothesis for the background noise was
right.
o Generated triple coincidence datasets for the r-Statistics analysis,
worked towards tuning the false rate of the triple coincidence analysis
to
maximize efficiency.
Weinstein:
- Worked on report on LDAS for DASWG
Yakushin:
1) Continue running simulations on the whole S2.
2) Preparing presentations for LSC meeting.
3) Trying to use Shourov's linear prediction filters with waveburst.
4) Regenerated all sky black hole merger MDC frames after Patrick Sutton
discovered that they were actually half sky.
5) On Erik's request working on generating MDC frames with events injected
not only at zero time lag but at time lags that are multiples of 5 s.
Lazzarini:
* Finalizing paper on optimal combination
of H1+H2 for use in stochastic searches;
* Started work with the calibration
team on the generation of time-domain h(t) frames. These will be a new
class of RDS frames,
RDS_Cvv (C for calibrated, vv for version of calibration). I am writing
a specification for these frames with the calibration team. Intent is to
have something to circulate by the LSC meeting. The calibration team is
preparing a presentation at the LSC meeting on this subject. Document
number is T040026 (not yet in DCC).
LDAS System Administration
--------------------------
Caltech
-------
(Dan Kozak)
* Continued to focus on getting data out of HPSS and into SAM-QFS.
* Worked on getting hpssmvr2 back online, first by working on it
myself,
then by trying to get the other HPSS support staff to work on
it,
both without success. Finally figured out a workaround
and continued
copying data.
* Dealt with an outage of the STK silo and worked with STK to resolve
it.
(Al Wilson)
* Tested NIS+ client with fedora. The RPM build that I have for RH9
worked just fine.
* New firmware installed on the switch in Millikan. This firmware
allows the
switch to work with copper Gig-E gbit. Tested, and this does
allow
these
adapters to work. But we still have the auto sense problem with
copper
Gig-E.
* Doing some HTML editing with the BB home page for CIT. Trying to
create a
one page overview of the systems at CIT.
* Replaced disk drive on datacache3 and 5. These systems are back to
full
working order.
(Stuart Anderson)
* Worked with Sun to fix memory problem in V880 at Caltech
* Worked with Foundry to hopefully solve a network switch
lockup problem in the large LDAS-CIT cluster.
MIT
---
(Keith Bayer)
* Ordering replacement harddrives for cluster machines.
* Getting dns entries for condor hosts.
* Replaced bad hdd in pcraid1 and rebuilt fs raid.
* Adjusted diskcacheAPI to read data shifted over to
pcraid1a/1b from other pcraids.
Livingston
----------
(Igor Yakushin)
* The power panel's unfortunate property to physically disconnect power
to LDAS in response to short outages has been fixed by Rusyl
and
Excel:
it is now powered through LDAS UPS.
* Requested IP addresses for grid computers.
Hanford
-------
(Greg Mendell)
* My LDAS sys admin work this week is a continuation of testing the
v1.0.0
release of LDAS with the createrds process, and setting up archiving
and
reducing of 1000 seconds of data every 100,000 seconds during
the
postS3
period.
(Ben Johnson)
* Completed most of the Condor harware/network setup. Including 4 new
DNS entries for the condor computers. ldas-grid and ldas-gridmon
have
been set up as has autofs on the nodes.
* Backup scripts have been successfully tested at LHO. Next step is
to
setup LLO to perform the backups, or perform any usage modifications
as
necessary.
* Continuing work on data corruption. We believe we're seeing the
smoking gun, but we'll try doing more analysis next week on
the DAQ
system itself to confirm our findings.
General Computing (Wallace)
MIT:
(Keith)
-Spec-ing out new gc mail server
-Spec-ing out new cad pc for Ken
-Added urops to gc:
Philip Richerme
Matt Abernathy
-Removed several old accounts from gc
Livingston:
(Shannon)
-Current bandwidth usage can be seen at
http://teche.ligo-la.caltech.edu/mrtg/LLO-Router/130.39.245.1_1.html
-Archived bandwidth usage can be seen at
http://teche.ligo-la.caltech.edu/mrtg/archive/ under the dated folder
for the week of interest.
-Finished moving all of the home directories to the new file server.
I began moving them manually and realized it was going to take too
long, so I wrote a script to move the directories and update the NIS+
tables. The script took 2 1/2 days to complete so I would have been
there forever moving the directories by hand. All of the home
directories are now on the RAID server. Still have a few
things to
finish up, but for the most part this is complete.
-Set up two laptops for Bonnie and Katrina to use for two tours this
week.
-Set up a replica NIS+ server to get ready to move more stuff off of
our old file server. Had some issues in the middle of this, since
the
UPS that baronne is plugged into had some problems during the
replication. It caused NIS+ to crash, then the machine was not
stable
and I had to reboot it. Commands like ps would hang and never
finish.
Luckily, the NIS+ tables were not corrupted. The last time this
happened it took us over 12 hours to get everything back. Once
I got
everything stable and running again with the replica server running,
I
decided to leave things alone for a while.
-Promoted the replica server to be the master server, removed baronne
as a server, and then created a new replica server for redundancy.
The
old NIS+ server is now a client. The mail server (abundance)
is now
the NIS+ server and touro is the replica.
-Installed a tape drive on abundance to back up the mail spools and
the system. Set up a script to run ufsdump nightly to the tape
drive.
Later this will be set up in a different fashion, but this will give
us
backups for now.
-Still working on compiling more apps and moving the old ones to the
new file server. Need to make sure that when I compile things
that
they do not link to libraries in /usr/local since not all machines
will
have the same packages.
-Set up another ntop server so that I can monitor bandwidth usage.
I
will later be setting up snort and ACID on the inside of the firewall.
The new Cisco 3550 works great for this and is much more flexible than
the old PIX. I can now span ports and use them for IDS purposes.
I
could not do this with the PIX. However, the PIX will be useful
when I
internalize the PCs.
-LSU contacted me about the old T1 router. I think it is safe
to say
that we will not need it anymore. I have e-mailed them and told
them
that I can return it since it is in their inventory. I also contacted
Ric Simmons and asked him to turn off the T1 lines permanently now
since the GigE has been stable for the most part.
-Put in a request for 4 DNS changes for LDAS. They are adding
4 new
gateways into the LDAS network. In my opinion this violates the
"spirit" of the LIGO IT security plan, but there is probably not much
that can be done about it. I count the number of gateways to
internal
networks to be at least 12 now, though I am sure there are a couple
I
am not thinking of off the top of my head. This also puts the
number
of operating systems at 4. Many of the gateway machines do not
have a
clear administrator responsible for the machines. The only layer
of
defense is the firewall and for some reasons this is not sufficient
at
this point. I am concerned that this will only get worse.
-Dell has provided a fix for the problems that I was experiencing on
the new Gigabit switch. I have not implemented it yet, but it
involves
a filter being put on our Foundry core switch. Hopefully it will
fix
the problem since I am only able to use 2/3 of the ports on the new
switch. So far I am more pleased with Dell's support than I am
with
their hardware.
Hanford:
(Christine)
- Purchased the yearly maintenance contract for Matlab.
- Put in a PR to buy 5 Sun blade 1500s.
- Notified users of a disk full problem, requested they clean up their
unnecessary files. I was also able to move some user directories
to
another disk to make more room.
- Over the weekend a disk containing user directories died. I
had spare
disks and everything was backed up to tape, so I was able to fully
recover in only 2 hours.
- Everyone received multiple emails containing the latest viruses.
Thanks to all the reminders and announcements from Caltech and Sysadmin
we were able to come through this latest round unscathed.
- Worked on a user's pcard problem. Something in the registry
or
software files has become corrupted. The pcard is looking for
a file
that doesn't exist and it won't run without it. Multiple reinstalls
of
the software have not fixed the problem. Still trying to fix
it.
- A user complained of multiple virus containing emails being sent
to
him via Caltech ITS servers. I contacted Caltech ITS, their spam/virus
filter was not catching that these emails contained a virus.
ITS agreed
to block the originating IP address, which came from an ISP in France,
and to notify the technical support at the ISP.
- Other misc. user support.
CIT:
(Mike)
-Worked on a few computers that were infected with viruses. This was
the
W32.Beagle.B@mm & W32.Netsky.D@mm, W32.Netsky.C@mm. I have cleaned
these
viruses with anti-virus removal tools, cleaned up the registries and
deleted other trails that were left behind from these viruses.
-I work on Calum Torries computer troubleshooting his solid works installation.
Every time he works on a drawing it ends up crashing. I have looked
for
viruses and came up with nothing. I then installed a current service
pack for
solid works and am testing this right now. This is looking more like
a video
card incompatibility issue with solid works.
-Backup all NTSRV's for an end of month backup. I also included the
PDMWorks
server.
-Loaded firewall software on Florence, Dot and Ryans computer.
I have a list
generated by Larry Wallace, on who is to have this software loaded.
I still
have many more users to load this software on.
-Moved some equipment around with Larry to make room to store additional
computers and monitors, down in the sub-basement.
-I removed a workstation from DCC that is being reloaded for Bob Taylor
over in
40 Meter. I am just about finished with this PC and will be swapping
out
Bob early next week.
-Swapped over Ruth Brambila & Easter Cunningham to the ITS network.
This was due
to an authentication problem that these users were having upon
logging in to the Business domain, located on the ITS network. This
was taking
anywhere from a œ hour to an hour just to login going through the LIGO
network. Plus there were many other issues they were having on the
LIGO network.
Everything seems to be running smoothly now.
-This is another week of many onsite and phone user support that
included
printing and software issues. I check out many computers that users
had
questions with on the latest round of viruses. The majority of our
users are
doing great job on being conscientious when receiving e-mail with attachments
from unknown users or even legitimate users with strange attachments.
Any
questions on strange e-mail, please do not hesitate to call General
Computing.
(Lisa)
- Spent most of the week getting the acrux replacement server tested
and ready
to go online. This included upgrading becrux so that the 2 servers
can use the
same spam database.
- Have started digging into the details of clamav, the virus software
that will
be running on the mailservers henceforth. Since clamav is an
open source
project, updates may occasionally include checking out the code from
source.
- Looked into using logrotate to manage the logs on the mailserver.
Unfortunately, it doesn't work. Am going to look into rottlog
next.
- Helped Larry chase down a couple of infected pc's.
(Veronica)
- CaJAGWR: Recorded the video of the last talk by Adam Burrows. Addressed
issues of remote participation from Hanford; at the end LHO decided
to go
with a dial-in and indicated they might be interested in a webcast
in the
future.
Compressed and posted the video of the talk. Adam had a lot of animated
simulations in his talk, and I am trying to compress them to add to
the
posting. Updated the seminars webpage and provided user support. Burned
the 2004 talks on DVD and sent off to LHO, as requested by Vernon
Sandberg.
- LSC web: Posting the MOUs/PRs as they keep (and keep and keep) arriving.
Updating the March meeting website. Working on the Contact Information
webpage of the collaboration.
Installed a webpage for motions to be discussed at the upcoming meeting.
- LIGO: Usual upkeep. Assisted Calum Torrie in choosing the navigation
for
a webpage on which he works. Provided a set of scripts for him to try.
Prepared high-resolution images per a request from Caltech Media Relations
office.
Updated seminars webpages.
(Larry)
-This has been an interesting week with the viruses that have been
going around.
We've had a number of machines that have been infected through a backdoor
on the
machine. They have all been taken care of.
As for the users they have been great. So far no major incidents from
someone
opening up an infected e-mail message. Many have contacted us(the sysadmin
group) concerning suspicious e-mail messages which is the correct thing
to do.
-Spent time with Shannon going over a number of items he is working
on at
Livingston. The NIS+ issue did work out in the long run but for a short
time it
looked like it was going to be a show stopper.
-Worked a number of printer issues. A number of them were paper jams
but others
appear to be the drivers on the computers needing to be reset in-order
for the
print to come out correctly. Still have a few drivers to setup for
the E2E
machines.
-Worked on a couple of the E2E servers to get them back into correct
working
order.
-Helped a couple of people out with licensing issues.
-Setup a second audio enabled webcam. Reprogrammed a couple of webcams.
Presently, we are doing some testing on the units. The one thing that
is a
problem is that so far we can only get the audio from the webcam to
work under
IE. This is something we will continue working on to try and find another
solution.
There will be one audio webcam setup for checkout. We are getting the
full
package put together and it should be ready by next week.
-Ordered a number of computers and related items. Working on get a
couple more
SUN orders out.
-Albert informed me that there is a new update for the VRVS but we
have not had
time to test it out yet.
-Worked a little on some of the engineering pkgs. being used. This
is something
that will probably need more support in the near future.
-A lot of regular user support.
SEI Structure Design/Fab contract:
ASI is applying fewer hours than expected. Charges to date are 33% under plan, including the additional 5% from rebaselining indirect, overhead and G&A. A re-plan is currently in process, to evaluate the impact of the recent slippage.
A payload configuration matrix has been developed for the latest range of BSC payloads. Finalization of the BSC configuration design is in procress. The design for the spring blade test fixture has been finalized; detail design is in process.
The Design Review meeting (for BSC & HAM structures combined) was
originally planned for March 16, but will now be late. Due to the amount
of detail involved, it is likely that a full day will be required for reviewing
each of the two structures. ASI is considering separating those and focusing
on the BSC to minimize the delay in the BSC structure, which is LIGO's
priority.
Actuator Testing and Redesign:
The four proposals for actuators received on 20 Feb were evaluated,
with 2 being judged as in the competitive range. Corrections were
made to our Design Requirements which were reissued to the 2 remaining proposers,
less the vacuum thermal performance requirements, which are being reworked.
These are expected to be issued by 3/5, with responses to be received by
3/15.
Position Sensor Probe Development:
The RGA scan of the 10 epoxy coupons representing 20 units of position
sensors measured an equivalent per unit outgassing
of 1.5E-13tl/s to 3.2E-12 tl/s
for the 5 flag hydrocarbon AMUs after subtracting
for the RGA head background reading, depending upon whether or not a 9/04
empty chamber reading was subtracted from the total. This range is to be
compared with the 7.5E-13 tl/s level allocated
for this item in T040001-00 (proposed levels for various components in
Advanced LIGO).
Seismometer Procurement:
This task is on hold pending management approval for SEI structure prototype
fabrication.
Galling/Dusting Test:
No further progress.
Preparing for the LSC, Suspension Summit and the LASTI test in June.
Having problems with IDEAS running a file of Dennis's.
Larry Wallace is providing support.
From: ctorrie <ctorrie@ligo.caltech.edu>
Advanced LIGO SUS
Prototypes for the Quad
Test mass
Parts for a single pendulum setup have been received from workshop
and will be assembled next week. The hope is to test interface and ergonomics
of the 40kg mass. Also it is hoped to test the use of
Test mass and Penultimate mass
Drawings for a double pendulum prototype are ongoing
Structure
A updated structure has been created that
has representations of non-suspended masses at the same heights as those
quoted in Norna's recent document T040028.
Its height is 2005mm and includes a section that will interface between
the structure and the suspension
Discussion are ongoing with
Layout
A 2-D layout representing the suspension has
been created in SolidWorks
AdvLIGO Coating Development
We had a coating teleconference with J.M.Mackowski.
Gregg, Sheila and Jim Hough also participated. One of the first coatings
on this new development program, a fused silica substrate coated with HfO2/SiO2
is being annealed at this time.
Jean-Marie mentioned that the absorption and scatter from
HfO2/SiO2 will always be higher than the Ta2O5/SiO2. The annealing
temperature for HfO2 needs to be lower.
The absorption is ~ 5-6 ppm; it can
be reduced by 1-2ppm with some tuning
but it will never reach the 0.5 ppm
level required for AdvLIGO.
Scatter is also higher, between 20-40 ppm.
This is due to the structure of the material's growth.
From: GariLynn
Billingsley <Billingsley_G@ligo.caltech.edu>
Subject: Good news from
Considerably lower absorption
was found in the sapphire piece from
http://www.ligo.caltech.edu/~gari/LIGOII/Downselect/LyonReportC040088-00.pdf
For additional information about this report, contact sanders@ligo.caltech.edu