Weekly Report for Week Ending May 22, 2003

The LIGO Executive Committee Agenda for Monday  May 26, 2003 will be:

 CANCELLED DUE TO MEMORIAL DAY

Special Announcements:

Weekly Report Highlights
 

LSC Issues (Weiss)

no report

LIGO I Construction/LIGO Laboratory Administration (Lindquist)

LIGO Weekly Site Telecon (Lindquist)

There was no site teleconference on Thursday, May 29, 2003.
The list of current actions revised to reflect the status of open actions assigned through May 15, 2003 may be found at ACTION LIST.

From: Ed Chargois <chargois_e@ligo.caltech.edu>

• Providing support to Caltech's Property Services Division, assisting with the annually campus Inventory.
• Provided assistance to the Detector Group (H. Armandula) with packing and shipping of two (2) Substrates valued at $1500 each to MIT (G. Harry).  Account Number P204317.
• Providing assistance to the LHO Site Manager (O. Matherny) with the research and procurement (with the assistance of G. Salone) of two (2) Electric Carts. Bids have been received from Taylor-Dunn,  E-Z-GO, and Club Car. This will be a GSA Purchase.

>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

Extracted forms that are currently exhibits in E960022-B and prepared them as stand-alone forms.  They were numbered and placed
in docspublic for access and will be posted on the web page when Veronica has more time next week.  Final edits were made to the main document reflecting the hyperlinks of these two forms as well as other minor changes within the document.

Began a cursory audit to compare electronic documents available but not showing up as available electronically through the web search tool.  Will try to compile the list along with when it was added for electronic access to determine the span of time when the linking between the web search and the electronic links broke down.  Further analyzed several documents in the archive directories that are not in the DCC database.  These will be a priority to pull and process so they are reflected in the database and through the web search tool.

Began work with George in walking through the current workflow and processes as he begins to analyze data relationships and possible areas in the current architecture that could be the cause of the instability presently experienced.

Laptop was returned all cleaned up and ready to try again.  Worked with Mike in final loading of programs and connections prior to the crash.

• Reminder to LIGO staff:  Please make sure that as a minimum, each document submitted has a LIGO number, author's name, and date.  It's surprising how many documents are missing one of more of these essential pieces of information!.

ACTIVITY

• No report this week.

From: Esther Cunningham <esther@ligo.caltech.edu>

Press here for ACCOUNTS PAYABLE HISTORY DATA .

From: "Brambila, Ruth" <Ruth.Brambila@caltech.edu>

• Completed an expenditure type change on NTD to change order 7.  Completed change order 4 to the subcontract to Excel Engineering 1051884, change order 4 to the subcontract to Excel Group 1055977.  Effected tax code changes from LA Sales tax to LA Use tax on 3 blanket po's to GP Manufacturing, Air Products, and Digikey.  Esther and I are working together to try to find an easier way to make such changes on the blanket purchase orders which get to be too long.  Change order #1 to CSRS has been completed, change order #7 to SIOM, change order #6 to Corning for 5 pieces of fused silica, change order #13 to Crystal Systems for 32 sapphire windows, the consulting purchase order for George Stokes, and change order #1 for CSRS have been completed.
• Placed and reconciled pcard orders as requested.  The monthly pcard credit limit for me has been increased from $50,000 to $75,000 after a transaction that exceeded my limit rejected.  This will allow me to continue to place pcard orders as needed without needing to worry about exceeding my monthly limit.
• Began entering change order revisions to the Access system for Purchasing Services.  Updating lobbying and debarment certifications in the database.

• This week I've been working with Ed Chargios on purchasing some carts for Handford.  We have received three quotes a decision is near.
• Also I've obtained a processing method on blades from Superior Jig for Janeen Romie.
• Processing PO's for the retrofit project..

• Provided information regarding Visitors for DCAA audit.
• Notified Donna Tomlinson that there appears to be overcharges for Medical and Dental coverage for a Visitor on the Visitor Program, she is following up to have a correction made.
• Continued close monitoring of activity in the Construction Award, current projection shows an over-commitment of $182K.
• Prepared and submitted 4 Cost Transfers to transfer costs for RFI Measurement Equipment from Construction to Operations.

 
• Financial reports can be found at: http://docuserv.ligo.caltech.edu/~fireport

From: irena@ligo.caltech.edu (Irena Petrac)

• AdLIGO RFP: The solicitation is in the works.  Addendum  No.4, providing additional technical clarification, was released to the potential proposers.
• REO: Change Order No. 16, for coating of additional LIGO substrates, is in review by the LIGO Optics group.
• Wave Precision: A Change Order NO. 23, for the purchase of additional optical flats, is in review by the LIGO Optics group.

• No report (vacation).

SUPPORT (Baldon, Lloyd, Tischler)

>Irene Baldon

• Processed the paper work for seven (7) new/revised trips.  There are approximately twenty-one (21) trips in various stages of completion before ticketing can be done and/or paper work completed.  Also assisted a number of P-Card holders with their travel plans/reservations or revisions.  Until further notice I will be doing travel for LHO travelers.
• The 2003 SURF Program will be held from June 16th thru August 22nd.  LIGO/Caltech will have twenty (22) students, LIGO/Hanford will have seven (7), and LIGO/Livingston will have three (3), with four (4) going to Pisa, Italy.  I have started ticketing the thirty-two (32) coming to one of the three (3) LIGO sites and Pisa will be handling the four (4) going there.
• Completed twelve (12) Expense Reports and there are nine (9) reports yet to be done.  I continue to contact travelers who have outstanding Expense Reports (more than one (1) month old) to ask for their cooperation in sending me their receipts so that they can be closed in a timely manner.  There  is only one (1) report more than 30 days old at this time.  Completed one (1) Expense Report for a traveler who has his own P-Card but was having difficulty and filled a Payment Request for him.
• Reconciled three 3) P-Card items which required numerous phone calls to hotels, car rentals, etc., in order to track down which traveler incurred which charge; and the pulling of each traveler's file to assure that charges are correct for each fee incurred.  Reconciled a total of seventy-eight (78) items for the month (4/15 to 5/15).  Assisted a few travelers with their questions/problems in regard to their P-Card and/or reconciliation.
• I'm happy to report that the new EFT Program for Travel Reimbursement Direct Deposit continues to be functioning well and although there have been a couple of minor bugs Michelle and I are working on them.  Prospects are good for a successful implementation.  If you have any questions or problems with EFT please let me know so that we can address it promptly.  For those of you who have not as yet sent me your form for participation please send it to me as soon as possible and I'll get you signed up immediately.  Thanks for your cooperation and help on this one.
• I continue to do MIT's travel to the sites for installation activities and also to assist them wherever possible.

• Processed the usual invoices for payment. Tracked and followed up on invoice problems. Reviewed and recorded payments processed by Esther the week of May 12.
• Processed the usual requisitions, change orders and payment requests.
• Continue to monitor contract and blanket order encumbrance and notify task managers when supplements are needed.
• Jim continues with data entry in the LIGO database and helping out in the DCC.

• F&B details for the NSF Review of AdvLIGO.
• Mayfield High School tour details.

Advanced LIGO (Frey)

No report this week...very busy trying to finalize the Advanced LIGO schedule and budgets to send to NSF Review Committee.

Project Web Site for posting schedule and progress related data continues to be updated with the latest and greatest.

• See http://www.ligo.caltech.edu/~tfrey/index.html

The Construction project quarterly report for the end of February was scheduled to be submitted at the end of March.  Irena has been working on this, but has been distrtacted by a number of other priority issues: LSC, contract activities for seismic isolation, optics coatings, etc.  We have discussed priorities, and the quarterly report is still slipping.

I have prepared and distributed another model for projecting FY 2009 operating costs and prepared a supporting document for Gary/Barry.

A response has been prepared and iterated for the DCAA audit questions.  A final response is being prepared by Caltech Administration.

The following change requests are pending or new.  Dennis Coyne submitted CR-030011 for the addition of an External Pre-Isolation (EPI) stage to the seismic isolation systems on the LIGO Livingston Observatories. This change is to improve the low frequency isolation of the currently installed seismic isolation systems as a retrofit with little or no disturbance to the alignment of the optics and without entry to the vacuum system.  I will put these change requests on the agenda for the next meeting of the Executuve Committee.
 

CR-030008	Furniture for the auditorium lobby, interaction area underneath the skylight atrium in the OSB, and on the second floor interaction area of the new laboratory/office building. (ON HOLD pending additional cost data.)	M. Coles	April 29, 2003
CR-030011	Seismic External Pre-Isolation at LIGO Livingston Observatory	D. Coyne	May 16, 2003

From: Cindy Akutagawa <cindy@ligo.caltech.edu>

• Various personnel/payroll /HR related work.

Quality/Safety (Tyler)

LIGO Hanford Observatory (LHO) and Interferometer Operations (Raab)

Brian O'Reilly was visiting from Livingston, working on S1 and S2 calibration studies, plus investigating tidal servo issues.  Bill Kells has been conducting a series of measurements having to do with understanding the performance of H1, in conjunction with improving the strain sensitivity. Details can be found in a series of elogs 5/19-21.  Robert Schofield reported on a series of acoustic mitigation experiments, starting here.

4K IFO

Full locks have been re-established on H1, after more than a week of difficulties.  Once again we do not have a smoking gun for why we could not lock, nor why locking is better now.  The x single arm locks had clearly degraded. but re-establishing these did not immediately yield full locks. Work continued on the 4k power increase.  WFS work resumed.

2K IFO, DAQ

Rolf Bork is visiting and testing LSC code on the 2k, endeavouring to rid it of crashes, and adding in new lock aquisition code from Matt Evans; see Rolf's elog.  This occupied the 2k for much of the week.  Earlier, comparisons of independent dc calibration schemes were made on H2, finding agreement at the few percent level between PZT fine actuator methods and sign-toggling/AC extrapolation methods.
 

LIGO Livingston Observatory (LLO) and Interferometer Operations (Coles)

no report

Detector/Technical Support (Coyne)

40 Meter Interferometer (Weinstein)

• SURF Matt Eichenfield from UNLV has begun work at the 40m. He's starting by working with Alexei on the MC WFS system. Welcome, Matt!
• Core digital suspension controllers (DSC), LSC, ASC, and other Electronics (B. Abbott, Heefner, Bork, Taylor):
• All installed optics (MC1, MC2, MC3, BS, ITMX, ITMY) are damping well with the new DSC system.
• Dennis and Conor completed the diagonalization of the sensor readings for MC1, MC2, MC3, starting with the P/Y biases they obtained from their recent alignment work. The results were reproducible, and close to what they were before we tore apart the DSC system. They will move on to diagonalizing the outputs (coils). They will then move on to diagonalize the core optics DSCs, and optimize the damping gains.
• Ben rearranged the end rack drawings to reduce the number of needed VME crates by one per end. The drawings have been reviewed by AJW and Jay, and Ben has begun to assemble all the hardware for the end racks (all of which should be in hand).
• Rolf reported last week that the work required to prepare the end DSC software should only take a few days. We are aiming towards completion of the system before Ben leaves for vacation in less than 2 weeks!
• Ben matched the gains and filters in the Generic DAQ Intfc Board in 1Y2 (new read-out for fast PSL and MC signals) to the requirements.
• Jay reports that the drawings for the core IFO LSC system, and for much of the ASC system, are almost complete; he just needs to add in some QPDs. He is contemplating designing a new board to interface QPD signals into an ICS110B for DAQ. He thinks that an LSC system can be assembled by July.
• Ben has updated the drawings for the 1X1 auxiliary rack (EOS's, mech shutters, video switch, illuminators), and has implemented the changes in the 1X1 wiring. Lisa is cabling up the cameras to the video switch. Dennis and Steve will order some monitors. Dennis and Lisa are preparing the EPICS code to control all these devices (c1aux).
• Alexei is preparing the EOS EPICS screens. We aim to complete this system in the next week or 2.
• We need to run AC power to the small racks that hold EOS electronics. Steve & Ben will pursue.
• Suspended mass mode cleaner (Ugolini, Miyakawa, Vass, Abbott):
• The MC1, MC2, and MC3 optics are all damping well, and the DSC's have been diagonalized.
• Dennis and Conor improved the alignment of the mirrors (looking at TEM00 fringes on the cameras) in preparation for acquiring lock with the new system.
• Lock acquisition has been problematical. The MC_F path to the PSL appears to be behaving normally, but the MCL path is not. Which makes sense, because only this path had changes made: Signal digitized by Pentek rather than ICS110B; New suspension code (one error in this code has already been fixed); Pole-zero module moved from a VME board to a digital filter; Dewhitening filter is now in use. Thus, the gain is substantially different, and it's not clear if all the filtering is appropriate. Dennis is taking transfer functions and trying to compare with ones made by Osamu before the DSC changes (unfortunately, Osamu is in Japan). Work in progress.
• Mode Cleaner WFS system (Ourjoumtsev, Abbott, Eichenfield):
• Alexei and Ben have been finding and fixing problems in the MC WFS heads (bad op amps replaced) and MC WFS demod boards in 1Y2 (bad op amps replaced).
• Alexei is fixing small problems in the (new, digital) WFS EPICS screens, and bringing over missing databases and screens from the sites so that the attenuators on the WFS heads can be switched. He's adding the DC WFS quad signal monitoring.
• Alexei now sees good RF signals on all 2*4 channels, as well as DC signals in EPICS. He continues to await good MC locking.
• SURF Matt Eichenfield is learning Matlab/simulink to model the MC WFS system properly.
• The output to the PZT steering mirrors on the PSL table are being debugged. We will pursue feedback to these PZT steering mirrors and to the MC suspended optics; so we are following the work done by Joe Kovalik and Valery Frolov at LLO.
• Suspended optics:
• Steve & Conor balanced and re-hung the PRM. It damps well, and is ready for installation.
• Steve & Conor, with help from Matt, balanced and glued the SRM. Bob is baking it, and it will be re-hung and prepared for installation by the end of this week. That's the last of the core optics to be prepared.
• PSL and ISS (O. Miyakawa, B. Abbott, R. Abbott, F. Nocera, D. Ugolini, S. Vass, C. Mow-Lowry): Working well. Development work on hold.
• Optical sensing (Smith, Ourjoumtsev, Goggin, Miyakawa):
• OPTICS PARTS LIST, Orders pending: 1 q-switch E/O shutter from Fast Pulse; 1 Brewster's angle polarizer for SPS anamorphic prism steering mirror from KLC; New mirror holders for PZT steering mirror.
• FARADAY ISOLATOR: Parts are in process of cleaning and baking by Bob. Vacuum-ready parts expected by 5/28.
• OPTICAL LEVER ASSY: Assembly on hold.
• IFO STEERING MIRRORS: The mirror holders were manufactured incorrectly and are being remade by the vendor. New parts are expected by 5/23. All other parts are in process of cleaning and baking by Bob. Vacuum-ready parts expected by 5/28.
• MODE MATCHING TELESCOPE: Parts are in process of cleaning and baking by Bob. Vacuum-ready parts expected by 5/28.
• SPS BREWSTER PRISM MIRROR: All parts, except for the Brewster prism, which is expected the week of 5/19 are in process of cleaning and baking by Bob. Vacuum-ready parts expected by 5/28; prism to follow.
• AUXILIARY VAC WIRING: Completed a spread-sheet detailing the in-vacuum wiring for auxiliary components inside the chambers: MMT focus motor, PZT steering mirrors, motorized beam flags. Gave information to Jay H. regarding the IFO ISC signal interface with the PZT mirror drivers. Bob has OSEM pigtails, vacuum-ready, for the MMT focus motor and motorized beam flags. The PZT steering mirror cables need to be RGA scanned, and a 25-pin connector added.
• Steve proposes that we move the mechanical shutters for MCR and SPS beamlines to the first table that they go through after exiting the vacuum envelope (AP1) instead of the table that they are sensed on; for safety purposes. Mike will look at re-layout of the drawing.
• Alexei is ready to install EOS's on MCR and SPS beamlines as soon as the 1X1 software is ready and AC power is run to the small racks that hold EOS electronics.
• Steve has almost all the parts that were designed by Larry Jones for our shutter flags (one goes in front of every test mass, to block or attenuate the beam if desired). Bob is vac-prepping the parts this weekend. Steve is ordering picomotor controllers for them. Mike will design beam dumps for them.
• Facilities and vacuum envelope (Ugolini, Vass, Jones):
• We plan to install almost all the remaining in-vac equipment for the first round of experiments, in the next month: PRM, SRM, ETMx, ETMy, Faraday Isolator, MMT, PZT sterring mirrors, all other steering mirrors, shutter flags, camera mirrors, optical viewports, etc. Mike will develop an initial alignment procedure. We hope to have this complete, sometime in July (Seiji arrives July 1 to help).
• Steve has LOTS of surplus vacuum equipment (including high-quality VAT right-angle manual valves) available to any interested LIGOn. Ed Chargois will ship some of it to the sites.
• Two PC's had virus warnings. They were decontaminated by Mike Pedraza.
• Steve ordered an A/C unit for the control room (which is starting to really warm up), and will call in electricians to wire it up.
• AJW ordered two new PCs for the hoardes of summer visitors.
• South Annex Bake Ovens (Taylor, Cardenas):
• Bob is currently baking the last optic (SRM) before final hanging.
• Bob has lots of clean/bake jobs from Mike (MMT, FI, etc); all six vacuum bake ovens, and the air bake oven, in the South Annex will be in use, this weekend!
• Bob will be building hybrid OSEMs for Janeen next week.

Thermal Noise Interferometer (Libbrecht)

Since our last report, we have focused on improving our noise curve at

high frequencies, above 1 kHz. By looking at the coherence between the

two arm cavities and by measuring the individual contributions from

various noise sources, we have determined that our noise floor is

dominated above 1 kHz by laser frequency noise and electronic noise.

We can reduce laser-frequency noise by increasing the high-frequency

gain of the mode-cleaner servo, and to do that we are designing a fast

Pockels-cell path to implement frequency stabilization up to 500 kHz.

(Our current bandwidth using the laser's pzt actuator is only 90 kHz.)

For this high-frequency loop to work, we have to have positive phase

margin up to at least 500 kHz, and higher would be better. Right now we

are in the process of finding and eliminating components that eat

phase above 100 kHz. So far, arrested culprits include an SR560 used as

a summing junction and a mini-circuits low-pass filter at the output of

the mixer, both of which contributed substantially to the loss of phase

at high frequencies. Some phase loss remains, however, and we must

identify and eliminate its cause before we can implement the

Pockels-cell path.

We can reduce the net contribution of electronic noise by turning the

power (or the discriminant) up and the electronic gain down, if we can

preserve the input-referred noise in the servo filter. Right now our

servo filters are at their minimum gains, and any external attenuation

would either raise the input-referred noise or lower the dynamic range.

We are working with CDS on both a new arm-cavity servo filter design

and new mode-cleaner electronics.

Separately, we have installed a broadband Pockels cell in the beam path

and begun to measure its response up to 100 kHz. Preliminary

measurements indicate a response of 0.02 Rad/V, very close to the

factory spec of 0.015 Rad/V.
 
 

LASTI (Zucker)

LASTI (Allen, Coyne, Mason, MacInnis,
Mittleman, Ottaway, Rollins, Shoemaker, Zucker)
---------------------------------------------------------------------
 

EPI Progress (Ottaway, Mittleman, Rankin, Zuo, Nayfeh, Allen)
----------------------------------------------------------------------
HEPI
We investigated the coupling between the platform motion and the
Streckhiesen ground monitor, and estimated the ground stiffness. Some
results can be seen on the LASTI ilog for May 20 and 21.
 

Jonathan installed the optical isolator on the pump controller and we
remeasured the hydraulic power spectra. Some results can be seen on the
Lasti ilog for May 20.
 

Myron has reshimmed pier #1, #2 and #4 (#3 had been done previously). He
has also been working on replacing the motor and pump. In doing this he
has found that the existing motor-pump combination are not as well
aligned as they should be, as specified by the manufacturer, and that
(as a result?) there has been a considerable amount of wear to the inner
seal.
 

MEPI
After the controller had been retuned last week we attempted to find the
optimum sensor correction coefficients. We did this by locking all 7
loops and measuring the transfer function from the ground to a witness
(either geophone on the optical table or streckheisen on the support
structure) as a function of the sensor correction coefficients. We are
currently analyzing the results.
 
 

LASTI PSL & other experiments (Ottaway, Rollins, Zucker)
------------------------------------------------------------
This week three of our UROP(Equivalent to Surf) students have started.
David Lopez will work with Gregg on looking through the S2 data to get
the test mass internal mode Qs and frequency. It is hoped that these
will provide a thermometer to measure the temperature of the optics and
give us a guage on the amount of absorption present.
David Kaplan will work with Dave and is developing a vacuum compatible
photo-diode for LASTI utilising the ultra quiet photo-diode circuitry
that Jamie has developed recently.
Dan Mason will work with Dave and is evaluating the CO2 laser that will
be used for adaptive thermal compensation for LIGO1
 

Work this week has mainly focussed on the adaptive thermal compensation
system and finishing off the paper on the sideband camera experiment.
 
 
 

ISC Periscope Design (Ken Mason)
------------------------------------------------------------
The ISC periscope is being redesigned to be stiffer and less sensitivity
to acoustical noise. Professor Samir is providing assistance in
designing an acoustical damper which will wrap around
the periscope.
 

Data Analysis and Computing (Lazzarini)

Simulation and Modeling (Bhawal)
E2E weekly Physics meeting
----------------------------
Matt, Hiro, Biplab discussed about ASC control and sensing matrix and some
needed modifications in SimLIGO PSL set-up after it was realized from e2e's
validation that some sideband terms contribute significatly in misalignment
signals (see below). The speed and compilation issues for code in the new
Linux boxes were also discussed.
 

Correct WFS signals
---------------------
(Biplab & Hiro) The validation effort for e2e's alignment signals finally
 produced something useful for the commissioners of the WFS system.
 We found the reason for the discrepancy between measurements and
 predication of WFS 3 and 4. The reason was that the theoretical calculation
 used hitherto (using Daniel-Nergis' Mathematica model) for non-resonant
 sideband (NRSB) demodulated signals at reflected ports took into
 consideration only those beating terms coming from plus-minus NRSB and
 carrier. The e2e run took into account other beating terms like those
 between Resonant sidebands (RRSB)  and "RRSB plus-minus NRSB" as well
 which turned out to be not negligible mainly due to the degeneracy of
 the recycling cavity.
 Biplab produced a new alignment matrix for the common modes to reflected
 port with demodulation by NRSB that will be compared by the WFS team.
 

SimLIGO
-----------
(Matt) This week I worked on the problem of ASC control in SimLIGO and at
 the sites.  The trouble is that the ASC sensing matrix is not diagonal
 and not easily diagonalized (due to noise and gain variation).  I have
 developed an algorithm for producing a robust control matrix.  The
 resulting control matrix gives stable control in SimLIGO and will
 (hopefully) be tested at LHO next week.
 

TAMA outreach
--------------
Dr. Hayakawa of ICRR started using e2e, under the guidance of
Prof.Ohashi. This is based on the MOU attachment regarding the
collaboration of simulation effort. They are going to purchase a
workstation for the simulation.  Hiro is helping to finalize their
specification. (Dr.Hayakawa is very impressed how Alfi is maintained
and updated.)
 

Code development and maintenance
---------------------------------
(Hiro) modeler code:
 The latest configuration has a erroneous setting to create a Makefile
 with inlining OFF, which makes things run much slower.
 For now, problematic Makefiles are modified manually to generate fast
 binary, but will be fixed soon so that inlining will be on by default.
 On the linux box, a few different sets of codes generate different
 results, although they supposed to generate same results. The cause of
 this problem is being traced.
 

(Melody)
        -  Preparing for the next build by modifying the build instructions
           to exclude obsolete programs and related software.
        -  Performed more benchmarks in the Linux
 

(Melody & Hiro) Speed optimization:
 Various optimization is being tested. Using -msse2 improves ~15%, -O(N)
 gives almost same results with N>=2.
 

(Hiro) CVS:
 The current CVS has many obsolete legacy contents and confusing
 documentation. The reorganization is discussed to remove those obsolete
 ones. Also, it was decided that Alfi will move out of the e2e CVS, but
 will be in a separate CVS.
 

(Ed Maros) Started process of removing old code.
 

Alfi
-----
(Bruce)
 - Added error viewer which was essential to Alfi use via Web Start (PR 400).
 - Completed work on port movement options (PR 392).
 - Working on matching this interface in the node External view (PR 380).
 

LIGO Data Analysis System
 

Software Systems (Blackburn)
The issue in the wrapperAPI which was preventing the same binary versions from
being run on both Redhat 7.3 and Redhat 9 has now been solved. This required a
"hack" of the code and relies on the operating systems properly cleaning up on
exit. A problem report was filed on this "hack" so that once we migrate away
from Redhat 7.3 altogether, we don't forget to undo the work around in favor of
proper exit code. This now paves the way for installing the final LDAS Beowulf
cluster with Redhat 9 instead of having to install several operating system
versions between now and the next science run on 400 nodes.
 

The memory leak in the metaDataAPI has been traced to swig names registered as
commands associated with delete_$class such as dbEasy and dbEasyStmt classes.
These leaks have been fixed and tested on the both standalone boxes and the
LDAS development system.
 

The mpiAPI can now distribute the wrapperAPI master process out onto the
Beowulf cluster. Previous versions of the code only allowed running the
master process on Node0 which occasionally caused a bottleneck in both the
CPU usage and available memory for the wrapperAPI master. Three new resource
variables in the mpiAPI's resource file allow/disallow and tune this new
functionality.
 

The bug which prevented the concateframe command from writing out frames
has now been fixed. A new bug was discovered with using the new RDS frames
which contained multiple IFO data in a single combined frame. This bug was
also quickly fixed. There is still an issue with using NDAS frames in LDAS
jobs. Part of the problem was traced to the frameAPI not handling an error
associated with bad NDAS frames and part of the problem is associated with
an as yet poorly understood corruption of some NDAS frames. An enhancement
to the FrVerify utility is being made to better isolate the form of the
corruption in the frame files.
 

The new createRDS command allows LLO and LHO channels to be integrated
into single output frames (not supported in previous releases of LDAS). It
will also have a new user option (-usertype) which will allow users to
uniquely distinguish "designer" RDS frames from the LSC standard RDS frames.
 

The S2 run RDS generation was completed last weekend. Early this week
the binning of the RDS frames into subdirectories containing 100,000 seconds
each was also completed. These frames are now available in the CIT archive
for LDAS jobs and replication to the Tier II centers.
 

One of the remaining memory leaks in the dataConditionAPI has been traced
to the the actions used in the slope dataPipelines. This should help to
pinpoint the location in the source code responsible for this small leak.
Work also began this week on an open problem report for mixed complex and
real vector algebra in the dataConditionAPI. The way the code is currently
written makes it difficult to fix this bug without major alterations, so
it probably won't be ready for the code freeze next week. A problem was
discovered with reading truncated ILWD files in the dataConditionAPI this
week. This was traced back to the ILWD library and is now fixed.
 

The migration of LDAS software to Redhat 9 using the newest GCC compiler,
version 3.3, continued this week. This is not the platform target for the
next LDAS release (0.7.0), but it is complex enough that we need to get it
started and understood as soon as possible. It has already uncovered that
much of the latest frameCPP code is not following the ANSI/ISO standard
close enough to work with the new GCC 3.3. It took several days to under-
stand the source of the problem, but it is now understood and fixed. The
fix is backwards compatible with the older GCC 3.2.x compilers, but they
didn't complain about the non-standard use of templates so the developer
needed to be up on the standard enough to not violate it but wasn't. It
is expected that this issue will occur in other modules of LDAS as we
continue to port the code to this newer compiler and OS.
 

Effort began this week to migrate our standard test suite into webpages which
will allow better tracking of new tests and test results in the future.
 

Hardware Systems (Anderson)
Caltech
-------
(Dan Kozak)
 

* Got permanent standalone QFS licenses for 4 more machines.  We now
  have two permanent standalone QFS licenses that we haven't activated
  (for framebuilders at the sites when they go to Sol9/SAM 4, presumably).
 

* Continued to monitor SAM-QFS behavior (staging and releasing) and run
  archiving by hand occasionally.
 

* Began research/preparations for re-archiving of all data to higher
  density drives.
 

* Helped with install of new Beowulf nodes in Synchrotron.
 
 

(Al Wilson)
 

* Tested the nodes (new machines). Will now build a master image for
  cloning. Also will compile a master list for RH9 rpm's from this
  configuration.
 

* Posted on ldas-sw the driver and procedures for setting up the network
  for RH7.3 installation for the new one rack unit servers. These
  machines come with e1000 NIC's and will not work with a base RH load.
 

* The new kernels are here. Updated the kernels on Linux boxes in the Dev
  system. The new kernel in 2.4.20-13.7 it is located at our RH mirror
  under updates.
 

* Put online (network connect) Primarydb2(10.16.10.1) and
  testdb2(10.16.10.2) That would mean that they are located behind the
  gateway server on dev. They have a standard load of RH7.3 and are
  configured as ldas class servers.
 

* Did some minor work on assy the new racks as physical therapy.
 
 

(Stuart Anderson)
 

* Continuing to support the facilities upgrade in the main LDAS computer
  room at Caltech.
 

* Successfully tested 8 of the new Beowulf nodes with a 60hr memory test.
 

* Received the last 2 STK9940B drives ordered for the central archive system.
 

* Received 2 additional 5000VA UPS units for the LDAS-CIT system.
 

* Upgraded the 5 ldasbox development machines to the latest RH7.3 kernel
  in an attempt to obtain increased stability--to early to tell yet.
 

* Investigated options for powering the new 10+TB SAM-QFS disk cache for
  the central data archive in Powell-Booth.
 
 
 

MIT
---
(Keith Bayer)
 

* Helping Scott Koranda with LDR setup on ldas.
 

* Put last pcraid into rack.
 

* Put datacon/beowulf into rack.
 

* Running memory burntest on datacon/beowulf.
 
 
 

Livingston
----------
(Igor Yakushin)
 

* u1d1 disk failed in t3-9. After replacing it, it could not get
  reconstructed: it turned out that there is also a problem with u1d5.
  t3-9 corresponds to /frame21 that contains S2 RDS data. I am currently
  backing up /frame21 to the new IDE RAID unit. SUN is shipping another
  disk to replace u1d5 which should arrive tomorrow.
 

* 70 new beowulf nodes arrived.
 

* Discussing with Allen, Rusyl and Shannon AC replacement and new power
  requirements for LDAS room.
 

* Got some of the ordered cables and UPSes from CDW.
 
 

(Shannon Roddy)
 

* Ordered some more PC interface cables for the KVM.  Igor needs these to
  hook up the new PC RAID.
 
 
 

Hanford
-------
(Greg Mendell)
 

* Multiple disk failures on t3-14 at LHO have made /frame20 on the LHO T3
  disk farm unavailable. S2 RDS data is kept on this file system. I
  rebuilt the LDAS diskcache at LHO without /frame20, so that LDAS can
  continue running without this file system.  However, /frame20 and S2 RDS
  data for the GPS interval [729273600 731529536) will be unavailable
  until we recover from the T3 problem and restore the data.
 
 

(Ben Johnson)
 

* Had difficulties restarting LHO LDAS Monday evening May 19, the
  Foundry FastIronIII switch froze, requiring it to be power cycled. Also,
  4 beowulf nodes encountered errors running "fsck" on themselves,
  requiring "fsck" ro be run manually.
 

* Had difficulties restarting LHO LDAS Tuesday morning May 20, an LDAS
  resource file had been changed, changing the startup port for
  diskcacheAPI. This unexpectedly caused the API to fail. Changing the
  resource file back to the original configuration allowed for a
  successful restart that afternoon.
 

* In beowulf node 32, replaced power supply and fan, newly supplied
  from ASA Computers. The node is presently being burntested.
 

* Working on two more T3 problems: replacement of a failed disk in t3-14
  and replacement of a controller card in t3-4
 

* The network driver for the new beowulf and datacon has been installed
  on both machines successfully. They are presently being burntested.
 

The final shipment of hardware for the large Caltech cluster is on track for delivery around the time of the
PAC -- on schedule with the information provided several weeks ago.
 

Data Analysis Activities (Lazzarini)
Shawhan:
* Checking the S1 inspiral upper limit analysis.
* Some more work on S2 data quality.
 

Yakushin;
1) After introducing noise rms estimate and new table columns into the
waveburst DSO, several new bugs appeared. Working to fix them.
 

Weinstein:
Weinstein:
- continuing to work on burst paper
- reading detector and stochastic papers
 

Creighton:
I have begun to implement an improved time-domain filtering routine in
LAL that maintains double-precision internally even when operating on
single-precision data, similar to the implementation in the
datacondAPI.
 

Reilly-K.:
I started moving the stochastic DSO functionality into
LAL routines, but then stopped because this may
not be the direction the group goes. I then started working
on moving functionality, which is now in the datacondAPI
to the LALApplySearch routine under LAL wrapper.
Whether we decide to use LAL apps or mpi processing
in the DSO, it will be necessary to have this functionality in
the apply search routine.
 

Mendell:
1) I have checked into LAL cvs under the lal/packages/tools/test
directory the initial code Malik Rakhmanov and I are writing to
independently check the beam pattern response functions in LAL. It is
called LALIndependentTestDetResponse.c. and is located in the same
location as the already existing LAL test code for the LAL DetResponse
functions, Dave Chin's LALTestDetResponse0.c. The basic code, based on
that written by Brian Cameron, compares the output from LAL for F_+ and
F_x with the expressions for these given in Jaranowski, Krolak, and
Schutz gr-qc/9804014.  The latter treats the detector as lying in a
plane tangent to a spherical Earth. Our tests to date find the
difference between the LAL routines and JKS are less than 6.0e-4. Malik
and I will expand the unit test run during "make check" when building
LAL to include some of our independent tests and to add comprehensive
independent tests for all detectors, quadrants of the sky, years, and
seasons that can be run from the command line. We will document this in
LAL.
 

2) Proposed to the PULG group that Mike Landry and I start writing a
stackslide DSO to run under LDAS.  A more complete proposal will be
given at the PULG face-to-face meeting in June.
 

Lazzarini:
Worked to formulate a X^2 analysis for the stochastic background analysis
analogous to the "X^2 cut^ used by the inspiral group. Looks promising.
Working with Romano to validate the approach. Basically, once we get an
upper limit, it would allow us to say the degree of to which the , e.g., time series behavior
is consistent with that upper limit . The analysis can be done by either time shifting the data
or asking whether the power distribution of the correlation kernel is consistent
with the underlying astrophysical process we are seeking to detect. It allows
us to use the broadband information of the interferometer in a manner not possible
with narrowband measurements.
 

Blackburn: Reviewing instrumentation paper.
 

General Computing (Wallace)
MIT:
(Keith)
-Supporting influx of urop students
-Hardware/network drops/software/accounts etc.
 

Livingston:
(Shannon)
-Worked on Marcel's computer to try and figure out why it was locking up.
Spent several hours swapping out various parts on a fresh install of
windows and it is still locking up after about 3 minutes.  It is not a
heat issue, so I am still not sure why it is locking up.  This machine
has already been back to the vendor and they can not find anything wrong
with it either.
-Set up a laptop for Bonnie so that she could do a presentation this past
Monday.  This will become a loaner laptop when she is done with it.
-Working with Larry all this week on a security audit.  I expect to be
busy with this through the end of the week.
-Ordered some more hard drives for the file server.  These will be
configured as a software RAID.
-Taking care of various pieces of paperwork and reservations for the
conference I am attending on June 8-14th.
 

Hanford:
(Christine)
- The first SURF student arrived.  I helped her get an account and get
setup on our computers.
- The new Cisco router has arrived.  I will be installing it next week.
- Had to cycle power on the old Cisco router due to memory errors.
- Delivered some documents and PC hard drives to the NSF.
- Lots of user support.
 

CIT:
(Mike)
Laptops:
-Finished up reloading Linda Turners laptop and configuring her user
environment.
-Started reloading a laptop for the loaner pool that was returned
from Hanford.
-Finished up loading a workstation for Greg McIvor who is a visitor up on
the third floor room 360 W/B. Who was using a older computer that I had
schedule for an upgrade.
-Started loading two additional workstations to replace a few older
computers. These replacements have much more processing power and will be
used for visitors/surf students that will be coming out to the project soon.
-Updated security patches on all NTSRV's and went through and search the
security logs looking for any intrusions. (All looks good)
-Transferred all NTSRV's ghost images to DVD; this will be an ongoing
process from here on out. After updating ghost images at the end of each
month I will also have a monthly backup copied to DVD. This will save disk
space on the server plus insures redundancy just in case of hard disk failure.
-This week I had a lot of onsite user/phone support that included printing,
software and OS issues. I also installed additional software for many users.
-Setup a meeting for the LIGO SEMINAR.
 

(Lisa)
- Built a solaris 9 jumpstart server.  I haven't tested it yet but I expect it
will take another day of tweaking to get the finish scripts right.
- Rebuilt bellatrix with solaris 9
- Rebuilt ascella with solaris 9
- Worked on a mathematica font problem for Hiro and Virginio
 

(Veronica)
- This week again time was spent almost exclusively on the preparation of
the Advanced LIGO website, finalising and testing. The website is now
posted at the LIGO homepage. I also prepared a CD version of the
website for the upcoming review and I am about to start burning the CDs.
- Project Science: prepared an email list of participants of previous
workshops and those who expressed interest. Sent out an announcement and
provided user support.
- LIGO website: Provided user support for ADvanced LIGO Suspension group
(webpage assistance and VRVS/VNC help). Monitored the VRVS of the last
LIGO seminar. Did usual upkeep.
- CaJAGWR website: user support; posted several updates.
 

(Larry)
-Spent a great deal of time reconciling my Pcard for the last month. Most of
the orders have arrived and the items have been distributed.
-Worked on a number of PC's and reworked a couple of network connections
in Bridge.
-Working on a computer audit with Shannon. So far things are going pretty good.
We should have a report out next week on what we have found.
-Misc. user support at the Observatory.
 

Advanced LIGO Development (Shoemaker)

For additional information about this report, contact sanders@ligo.caltech.edu