Weekly Report for Week Ending October 13, 2005

The LIGO Executive Committee Agenda for Monday October 17, 2005 will be:

(Meeting time: 10:30 am Pacific Time)

Announcements

Comments on Weekly Report

LSC Issues (Saulson)

LIGO Lab Operations

  • Administration (Lindquist)
  • Sites (Raab, Zucker, Shoemaker)
  • Commissioning (Fritschel), Detector (Coyne)
  • Campus Research Facilities

40-Meter (Weinstein)

TNI (Libbrecht)

LASTI (Shoemaker)

  • Data Analysis and Computing (Lazzarini)

R&D and Advanced LIGO (Shoemaker)

CHANGE CONTROL BOARD/TECHNICAL REVIEW BOARD SESSION AS NEEDED

  • No open change requests

Special Items: Status of Advanced LIGO Budgets for FY 2006.

Special Announcements:

Weekly Report Highlights:  Last night, after much work by Rob, Rana, Matt, and Osamu, Osamu locked the 40m in the full AdLIGO configuration.  He repeated it, 4 times. Longest lock was ~ 20 minutes.  See below.

LSC Issues (Saulson)

No report

LIGO Laboratory Administration (Lindquist)

STATUS OF LSC MOUs (Lloyd)

  • No report.

Non-LSC MOUs (Lloyd)

  • No report.

SITE TELECONFERENCE (Lindquist)

  • A site teleconference was held Thursday, October 13, 2005.  The following issues were discussed:
  • FY 2006 Budgets – Site budgets seem to be acceptable within $20,000 magnitudes.  There continue to be uncertainties with regard to budgets for commissioning support and travel in support of science running.  MIT requests have been received and will be entered into the budget model.  The primary open issues relate to Advanced LIGO and Advanced R&D needs.  We will meet tomorrow to review the current status.  The status will be queried during the Executive Committee meeting on Monday.
  • Property Issues – The LDAS Computers are on site in Livingston.  Some shipment damage was experienced and is being reported.  Livingston is in the process of firing them up and checking them out.
  • The list of assigned actions updated through September 01, 2005 will be found Here.

PROPERTY MANAGEMENT (Chargois)

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

  • Nothing significant to report.

DOCUMENT CONTROL CENTER (Turner, Mak)

>From: Linda Turner - turner@ligo.caltech.edu>

Nothing significant to report...same projects as highlighted last week.

>From: Cleveland Mak <mak_c@ligo.caltech.edu>

  • Assisted in sending out three contracts ("Advanced LIGO Seismic Isolation System", "Document Control and Management System", and "Seismic Isolation System Miscellaneous and Wall Components").
  • Assisted in sending out the RFQ for "Fabrication for Vacuum Housings for the Pod Assemblies" to 2 additional companies.
  • Processed and distributed 2 DCN's.
  • **Scanning Project Update** - Progress continues in the preparation, organization and scanning of of Larry Jones' old files..
  • Activity:

Week Ending

10/13/05

In

Out

Packages

26

12

Faxes

31

10

 

FINANCIAL SYSTEMS (Cronin, Brambila, Kaufman)

>From: "Cronin, Holly" <Holly.Cronin@caltech.edu>

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

  • There is a TechMart training session scheduled for October 26, 2005 through the main Procurement office for many LIGO individuals. If you would like to attend a TechMart training session and have not yet signed up, please contact Monica Marquez at ext. 2566 for scheduling. There is a new website that provides on-line training at http://www.atc.caltech.edu/finance/PPS/techmart/ which may provide useful information and supplement your training.
  • Provided basic training in preparing a requisition and approving it in TechMart for two LIGO individuals so that a new order could be submitted and approved. The requisition was submitted, approved, and the order has been placed and the bulk of the order has shipped.
  • Finishing up the reconciliation process on the billed credit card orders for October 2005 in preparation for the reconciliation due next week. There are only two pending invoices, so far, that I am waiting for responses to complete reconciliation.
  • Completed two computer maintenance orders issued to Dynamic Systems for the first year of a 5-year agreement for LIGO hardware and software maintenance support located here at Caltech.
  • Working on completing change order #1 to the JPL work order for Helena so that additional work can be performed by JPL.
  • Working on finishing the remaining four no-cost extensions for Garilynn.

>From: Florence Kaufman <fkaufman>

SUBCONTRACTS MANAGEMENT (Jasnow, Salone)

>From: Gina Salone <gsalone@ligo.caltech.edu>

  • Nothing significant to report.

>From: Ed Jasnow <jasnow@ligo.caltech.edu>

  • We are still waiting for the receipt of the signed contract from Cangelosi-Ward contract for the construction of the LLO Science Education Center.  The contractor sent it out earlier this week.
  • The architect for the Livingston SEC is preparing a proposal for the performance of construction management on the project.

SUPPORT (Baldon, Hiroto, Lloyd)

>Irene Baldon

  • No report.

>Julie Hiroto

  • No report.

>Dorothy Lloyd

  • No report (sick).
  • Jim continued with data entry in the LIGO database and helping out in the DCC.

PROPOSALS and REPORTS (Lindquist)

The Request for supplemental funding for the Outreach Building in Livingston has been enhanced one more time and submitted to the NSF.

DCC Steering Committee (Lindquist)

The purchase order with Synergy for the new document management system and related hardware requisitions are in process.

CHANGE CONTROL/CONTINGENCY MANAGEMENT (Lindquist)

  • No open change requests.

HUMAN RESOURCES (Akutagawa)

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

  • No special activities to report.  The Staffing Committee meeting is scheduled for October 24, 2005.

Quality/Safety (Tyler)

>From: Bill Tyler tyler@ligo.caltech.edu

Nothing significant to report this week.

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

Summary of Commissioning Activities at LIGO Hanford Observatory (compiled by M. Landry)

M7 ran last weekend with mixed success.  H1 (H2) ran with ~65% (~60%) up time, or ~50% (~43%) of the run in science mode.  Lock stretches were short on H1, often clipped at several hours, and falling shy of the S5 range target by several Mpc.  The IFO performance was sufficient however to test DMT monitors, the mandate for the run, and new codes were exercised and reviewed (see the elog Oct 7-8).

We would like to welcome Malik Rakhmanov of Penn State, who has moved to Richland for an extended stay at LHO.

Commissioning highlights are bulleted below:

4K IFO

  • Recall from previous weeklies that ASPD saturation and AS_I latching issues were hampering H1 performance.  As a test, it was decided that we should swap back in the old ASPD4 photodiode and see if the situation improved.  Indeed the latching problem has gone, however, we still have saturations that last for a fraction of a second that drop the range by many Mpc.  So, what is the difference since the 11.6 Mpc days of September?  One change appears to be that AS_I sits atop a larger DC offset than before.
  • Re: 1Hz lines and subharmonics.  We ran with the 1PPS ramp off during M7: the glitch previously observed in the error signal went away but the lines in the control signal remained present.  We look forward to the timing groups sinusoidal replacement to the ramp signal.
  • a report on the recent REFL beam stabilization efforts is given here

2K IFO

  • saving both locks and photodiodes through better triggering: new LSC code was loaded to trigger the AS fast shutter and switch back to ASPD5.  The code was problematic and backed out for modification.
  • ASPD4 modifications were made

DAQ

  • many H1 and H2 CDS processors had new code loaded on the Monday bootfest, including modifications to power sequencers

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

The MOPA laser power started fluctuating erratically by 15 percent last Thursday causing problems for the interferometer. After cleaning the laser cooling lines the power fluctuations went away but the beam alignment into the power amplifier was changed, either during the cleaning or as a result of temperature change of the master oscillator cooling plate. Re-alignment of the beam into the power amplifier mostly restored the output power (85% of the pre-cleaning level).

Another side effect of the cooling line cleaning was excess acoustic noise in the PSL enclosure from the air trapped in the chiller hose. The coupling for this noise appears to be surprisingly large, so that it appears in the interferometer dark port spectrum. The air was driven out by re-connecting the chiller lines and the acoustic peaks were reduced below the interferometer noise floor.

The alignment bias modules were upgraded in order to reduce the output noise and provide better protection against power supply failures. Unfortunately, coincident with this upgrade a new noise contribution appeared in the frequency range between 50 and 200 Hz, which is consistent with a pinkish force noise source (spectrum in displacement falls like f**-5/2 and is essentially stationary). Investigation of this new noise source is underway in parallel with efforts to improve frequency noise rejection at high frequencies.

Some other excess intermittent noise in one of the anti-symmetric port photodetectors was tracked to a missing solder connection on the shield of the RF cable.

The reduction of the low frequency (below 1 Hz) relative motion of HAM1 and 2 was demonstrated by adjusting the angular HEPI correction on HAM2.

CDS (Bogue)

  • Installed new state vector code with the help of Dave Barker.
  • Helped Sam W. with getting the right channels to support his REFLPD work.  This is not completely resolved yet.
  • Built a script/ssh server for cds.  I have not yet migrated the accounts and scripts from other boxes to this one yet, but the box is installed and available.
  • Worked more on the new raids.  There are some scsi problems that need to be worked out.
  • Restored data from the backup frame builder after ldas had a hardware failure.
  • Cleaned and organized the spare parts cabinet in the mass storage room.

CDS Code Support (Khan)

Ash is on vacation this week.

Education and Outreach (MZ for Thacker)

John is at the American Science-Technology Centers annual meeting in Richmond, VA waving the LIGO Science Education Center flag.

Site Safety and Security (Riesen)

Most of this reporting period was consumed with the safety/security system software change over.  The site is now using the new software and all functions are operational.  There was some work-around needed due to the fact the new software has a bug and a new "bug-less" version will be available around the first of next month.

Updated and posted the Laser User Authorization signage.

Found no laser or site safety concerns during my weekly site walk thru.

LLO General Computing (D. Giardina)

  • Kantec software upgrade:  had a few problems with the new version that took a couple days of trial and error testing possible solutions.  Situation temporarily resolved.
  • Assisted Bonnie with more html documents.
  • Norton will not install on one laptop.  Situation was reported this morning. I am looking at it today.  Will advise when solved.

AdL SUS/SE Mechanical Engineering (Spjeld)

Quad SUS Installation Fixtures

  • redesigned lift table support brace, lift arm, transport table support frame
  • meeting with Calum and Ken Mailand regarding redesign and manufacturing plan of fixtures
  • updating additional parts and drawings in progress
  • uploading revised files to PDM Works vault in progress

LDAS/Condor and data analysis (Yakushin)

Storage/Condor/LDAS admin:

1) The 140 new nodes are installed. Working on changing IP addresses and various configuration files. At least 4 nodes have obvious disk problems and would not boot. The others at first glance are running but I expect badblocks to find more bad drives.

2) The temporary HVAC (10 tons extra to whatever the building's handicapped HVAC can currently provide) so far was able to handle the load (all 210 nodes). However, the generated heat would increase as the new nodes start executing jobs. The estimated repair time for the building's HVAC is 2 weeks.

3) T3-19 went down on Saturday night. Powercycling it on Sunday morning brought it back to life. However, syslog showed memory problems with the controller. The controller in T3-19 was replaced during Tuesday's fb1 downtime.

4) It turned out that the upper controller in 3510 (that was replaced by SUN a week ago) never actually came up. I would need another fb1 downtime to either try to reseat it or replace it again.

5) Got quotes for service contract for the old 15 ton unit from Star.

Data analysis:

1) Revisited LIGO-GEO SG21 simulation triggers to understand the problems that Siong found in hrss(H1) vs hrss(H2) plot (besides the expected diagonal line, there was and extra line at a different angle).  It turned out that that line was due to the injections that were outside of waveburst frequency band.

2) Studied a set of LIGO-GEO production triggers in which GEO detector is shifted by 17 seconds in addition to the standard set of time shifts done by waveburst. Found a problem in the way I did that shift.  Corrected the bug and regenerating the triggers now.

Initial LIGO Detector Science & Engineering (Coyne)

No report.

40 Meter Interferometer (Weinstein)

There was a meeting of the 40 Meter Technical Advisory Committee (by telecon) on Thursday, 10/13, at 8:30 am Pacific. We reviewed the status of lock acquisition, e2e simulations, and DC readout design and procurement. Slides here (pdf).

Nergis reviewed a proposal to inject squeezed vacuum into the 40m interferometer, sometime next spring. Slides here (ppt) and proposal here (pdf).

IFO Commissioning

  • There was steady progress towards full lock of the 40m, with Rana, Rob and Matt (RRM) working at night, Osamu during the day.
  • RRM implemented the analog common mode servo (using the REFL DC signal) to the PSL AOM (the MCL path), with a UGF @ ~3 kHz, (still being tweaked). They got a factor 10 rejection of noise at ~ 200 Hz. The only CARM actuation on the ETMs is a slow script servo to keep the analog MCF error signal close to zero.
  • They then were able to reduce the CARM offset, to the point where the power was ~ 80% of the way to full arm resonance. That brought them "over the hump" of the REFL RF (166I) signal; they saw the signal rise and then fall towards zero as they reduced the CARM offset. So, when they manage to smoothly hand off the signal from REFL DC to REFL RF, we're pretty sure it will bring the ifo into full resonance. But it's tricky to hand off the signals in the analog servo (would be easy with the digital servos). Much of this success was due to careful loop filter shaping.
  • RRM deliberately misaligned the arm cavity so that any beam jitter would couple more strongly to the arm cavity error signal. But they did not see any change in the noise above 50 Hz. They conclude that beam jitter (eg, from noisy in-vac PZT steering mirrors) is not a significant contributor to the noise aboe 50 Hz. The PZT mirrors could still be introducing significant frequency noise, present on the arm cavity error signal but not the mode cleaner error signal (the mirrors are in between).
  • Osamu, meanwhile, was using the POX+POY RF signals to control CARM and reduce the offset. Switching from TRX+TRY DC signals to POX+POY RF is now easy and routine. After much careful loop shaping, he was able to also "get over the hump". Late Thursday night, he at last brought the IFO into full resonance, ie, full AdLIGO configuration. He was able to hold lock for many minutes, and regain lock when lost. Arm light power was ~ 78 (arb units), very close to our expectations given the losses in the arms. He took transfer functions (clear RSE peak, no sign of an optical spring peak) and noise spectra. Hooray!

IFO Modeling

  • At the 40m TAC meeting, Ken emphasized the importance of reliable e2e modeling to address a variety of important issues in AdLIGO.

DC Detection Development

  • Mike has completed the part drawings for the output mode cleaner, and awaits a decision on the material. We want a well-damped material, so that the PZT driving does not excite the mode cleaner body resonances too much. We were considering brass, but Dennis, Dave Tanner and other TAC members cautioned us against it, due to poor high-vacuum properties. Seems like copper is the best choice, so we are now investigating availabiliy, cost, and machining issues.
  • Rob and Rana are thinking about the design of the DC PD electronics, preparing for potentially large power (up to 250 mW).
  • Rick and Betsy at LHO have sent us two REO supermirrors with T=1.4%, and five HR mirrors (at 43.5degrees incidence, P-pol) for our output mode cleaner. We need the HR mirror to work at ~ 10 degrees, P-pol. Helena and company will test the HR mirrors at different angles; they may work quite well. Along with a curved mirror from Peter King, this means we have all the mirrors we need for the OMC.

Electronics, Controls

  • The ETMY oplev laser diode died. Dan changed it out and got the oplev servo working again.
  • Dan and Ben are preparing the hardware and software to at last install RevB coil drivers on the three mode cleaner suspensions. These have separate bias and drive paths, so the diagonalization procedure will not throw off the alignment of the mode cleaner.
  • In preparation for moving the MCR RF PD to greatly reduce the beamline length (and therefore the potential for misalignment and clipping), Bob put in new RD PD cables on the AP table.
  • Ben has delivered new RFPDs for the sites, and is now preparing 4 new ones for us.
  • Ben continues to develop the design for his fast meachanical shutters.

Lab Infrastructure

Thermal Noise Interferometer (Black)

No report.

LASTI (Ottaway)

No report.

Data Analysis and Computing (Lazzarini)

Simulation and Modeling (Bhawal)

Weekly E2E Physics Meeting

Sany Yoshida presented some preliminary results from the e2e model of seismic isolation system for AdvLIGO and comparison with Matlab model from Stanford (details below). The mathematical formulation of Modal Model was also discussed.

AdvLIGO Scattering Noise

(Hiro) Together with Mike Smith, using twiddle and analytic formulation, the transfer function corresponding to re-injection from various ports back into Core optics is calculated - to be completed soon.

AdvLIGO Seismic Isolation

(Sany Yoshida) Made some progress in AdvLIGO seismic isolation modeling. We found an error in the ABCD (state space) matrix that we had created based on the Stanford group's model. After correcting the problem, we tested the matrix with our e2e code by giving an impulse input to the ground X motion input port (one of the 87 input ports). The time series of the response from the stage1-X (the X motion of the first stage of seismic isolation) output port is identical to the impulse response obtained by the Stanford group with their Matlab code. We still have the problem of having unexpected output from other output ports. The problem is currently being investigated.

AdvLIGO Diffraction Loss

(Biplab) Had discussion with Juri Agresti who is calculating diffraction losses using Huygens-Fresnel integral approach -- basically the approach followed in Ray Beausoleil's code. We compared some of his preliminary results with my FFT results and they are in agreement with each other.

Will complete a LIGO note on AdvLIGO diffraction losses by tomorrow.

Modeler

(Hiro) 'dual_sum' module is completed, and now being debugged.

(Melody) Figuring out and documenting Matt's modeler scripts for remote processing.

Alfi

(Bruce) Completing work on Alfi node validation and node validation in regards to how it affects the quitting of Alfi.(Problem Report 495)

Data Analysis Activities (Lazzarini)

Shawhan:

  • Helped finalize S3 untriggered burst search paper for Amaldi proceedings.
  • Followed up time-shifted coincidences in the S4 untriggered burst search.
  • Did hardware signal injections during the M7 run.  Also worked with Vuk Mandic on modifications for S5.
  • Modified conlog and gwistat code to handle the new interferometer state vector information, and cleaned up conlogger channel lists.

Yakushin:

  • Revisited LIGO-GEO SG21 simulation triggers to understand the problems that Siong found in hrss(H1) vs hrss(H2) plot (besides the expected diagonal line, there was and extra line at a different angle). It turned out that that line was due to the injections that were outside of waveburst frequency band.
  • Studied a set of LIGO-GEO production triggers in which GEO detector is shifted by 17 seconds in addition to the standard set of time shifts done by waveburst. Found a problem in the way I did that shift. Corrected the bug and regenerating the triggers now.

Mendell:

  • My top priority between now and the start of S5 is to write driver scripts for generating SFTs during S5.
  • In addition, I continue to make progress towards producing S4 results for the November LSC meeting using the StackSlide code. The StackSlide review F2F was held successfully at Caltech on Sept. 28 and 29 with Mike Landry, Peter Shawhan, Teviet Creighton, Phil Willems and myself attending. Several action items coming out of that meeting will be addressed in the next month by myself and the pulsar group.

Mandic:

  • I have calculated the coherence between H1 and L1 DARMERR channels for all of S4 with 1 mHz resolution. The 1 Hz harmonics are very clear in the coherence, but the 1/4 Hz harmonics are not. A couple of simulated pulsars lines could also be observed in the coherence.
  • I also ran the stochastic analysis for the H1L1 pair with 1/32 Hz resolution. The sensitivity (theoretical sigma) obtained in this way is similar to previous estimates (using 1/4 Hz resolution). Removing the 1 Hz harmonics changes the point estimate from about -1*sigma to +0.5*sigma.
  • I have also submitted the paper studying the accessibility of the pre-Big-Bang models to LIGO to Phys. Rev. D and to the archive.

Other:

LIGO Data Analysis System

Software Systems (Blackburn)

LDAS

Added the functionality in the contolMonitorAPI client to pop up a text box for entering username and password when the use of an X.509 certificate isn't fully supported yet. This will be removed shortly as the fix for using host certificates to authenticate between two services running on the same box has been solved in TclGlobus.

The managerAPI has been tested against all ldas commands using X.509 certificates to submit the jobs via a custom test client. Peter Shawhan has been contacted about the availability of this test client to use as a template for incorporating TclGlobus X.509 authentication and job submission into GUILD and LIGOTools.

The ldas-ps/ldas-kill commands have been modified to robustly handle some ssh key problems (PR#2908). Along with these changes, the commands have been modified to search which APIs are to be running and only gets the host names for those APIs (PR#2903). Several of the packages in /ldcg had pkgbuilder corrections to their rules. These include cvsgraph (PR2925), octave (PR#2757), lam (PR#2858), tkcvs (PR#2928), rxvt (PR#2929), tkdiff (PR#2930), and lsof (PR#2927). The solaris installation of tclglobus was also redone after it was discovered that a patch needed to be applied for use with Solaris 10.

Efforts continue in understanding why some frame files do not get renamed from .tmp to .gwf (PR#2900). It appears as if this condition is isolated to the concatFrameData command.

Continuing to wait for the Solaris 10 and Fedora Core 4 upgrades at the sites to take place before releasing the next version of LDAS. Looks like early next week is the soonest this will take place.

TCLGLOBUS

Resolved the issue when running XIO GSI clients against XIO GSI server on Solaris 10. The problem was that IP to host lookup was not providing a fully qualified hostname on the LDAS-SUNDEV1 workstation. This was found to be due to the ordering of services to use in this request in the nsswitch.conf file. It was using nis+ before using dns. The nis+ did provide fully qualified names, while dns does. Globus is very sensitive to this and will not authenticate if fully qualified names are not provided for reverse IP lookups.

Resolved the problem with using two XIO GSI servers communitating with each other using only host certificates whent he services were running on the same server box. This will be needed by LDAS to allow the controlMonitorAPI server to communicate with the managerAPI.

OSG/GRIPHYN/IVDGL

Successfully executed the binary inspiral work flows on the University of Wisconsin, Milwaukee and the Penn State University LIGO OSG Production clusters. This did however require the costum patches to VDS to be installed over the official 0.2.1 release of OSG. This also is the first demonstration of running on a non Condor job scheduler - PBS (Portable Batch Scheduler) at PSU.

Continued to work with Murali at PSU to configure the voms-proxy-init configuration. Murali has now made the change and testing will begin next week.

Performed weekly yum update and reboot of the nodes in the OSG-ITB cluster.

Reported to Paul Avery LIGO's historical experiences developing a LIGO application to run on the Open Science Grid. This resonates with reports from other groups on the difficulties of making progress in a manpower limited activity driven primarily by volunteers.

Drafted a plan to the OSG IEB outlining participation LIGO and the LSC have planned for utilizing the OSG Grid over the next 5 years. The draft has been circulated to the LSC Computer Committee and is expected to be delivered to the IEB at the end of this week.

Recieved a request from the public relations officer for OSG to provide a picture and 130 words describing LIGO's use of the OSG for scientific analysis for a brochure to be made available at this years Super Computing Conference in November. Duncan Brown has agreed to  construct the needed material  based on the binary inspiral search.

Hardware Systems (Anderson)

Caltech

(Dan Kozak)

  • Helped with 3510 crash/SAM metadata problem at LLO.
  •  Helped with T3 crash (/frames data problem) at LLO.
  • Preserved fb0_frames data for CDS at LLO.
  • Setup SAM-QFS for M7 at LHO.
  • Repacked M7 data strewn across many tapes (will finish today) at LHO.
  • Worked with Ben on several occurrences of bad md5 values in LDR database at LHO.
  • Worked on dataserver-cit problem that caused reboot when a file in /archive was stat()'d. Recovered, samfsck'd and reported to Sun.

(Phil Ehrens)

  • Updated m27 to Fedora Core 4 and added an additional 512Mb of rambus 800 memory to it.
  • Discussed RPM package building and management with Erik Espinoza.
  • Worked with Mary Lei and Michael Samidi to get tandem2, tandem3, and tandem5 usable for TclGlobus development.
  • Created list of all machines on 6th Floor Millikan with primary user, OS version, and GC IP address.
  • Updated all desktop machines on 6th floor with Erik Espinoza's Sendmail configurations, Octave, and a few other packages.
  • Tried and failed to diagnose (twice) mysterious hard lockup of the machine ldas-suntest3. This machine drops all ports and video sync without warning.
  • Did full yum update plus install of Octave on all desktop FC4 machines.
  • Issued and installed host and datarobot certs for ldas.mit.edu which had expired due to expiration email going to Hari Pulapaka or Naveen, or ??.
  • Changed nsswitch.conf host order to 'files dns nisplus' because nisplus was returning the unqualified hostname which was inconsistent with the globus library host name lookup, and was therefore preventing correct functioning of TclGlobus on Sun machines with nisplus active.
  • Upgraded dziban to Solaris 10 with mirrored 80Gb drives. Installed Apache and Subversion. Investigating installation of "viewsvn", a php based server-side subversion repository browser.

(Erik Espinoza)

  • Contacted 2 vendors for budgetary quotes.
  • Configured system in cluster to send e-mail out.
  • Configured basic kickstart setup.
  • Spoke with other cluster admins and vendors about IPMI (more specifically Serial over LAN woes).
  • Researching issues with 64 bit linux on v40z.
  • Finalized ligoconf, nis-utils, & pam_unix2 rpms. Tested on x86_64 & i386.

(Stuart Anderson)

  • Researching component prices for next CIT cluster to optimize the configuration.
  • Collecting initial budgetary quotes for the next CIT cluster.
  • Still shuffling cluster computer equipment between Lab sites for S5.
  • Finalized the new L4 RDS frame configuration for S5.

MIT

(Keith Bayer)

  • Received 7 blade Foundry switch from LLO.
  • Troubleshooting condor job eviction problem.
  • Added new condor users to ldas-grid.
  • Rebooted cluster node 49 after fs on hda2 reported errors and switched into readonly mode.

(Junwei Cao)

  • LDRdataFind server is started at MIT.
  • Meta data in the LDR database is updated to be usable from MIT cluster.

Livingston

(Igor Yakushin)

  • The new 140 nodes are installed. Working on changing IP addresses and various configuration files. At least 4 nodes have obvious disk problems and would not boot. The others at first glance are running but I expect badblocks to find more bad drives.
  • The temporary HVAC (10 tons extra to whatever the building's handicapped HVAC can currently provide) so far was able to handle the load (all 210 nodes). However, the generated heat would increase as the new nodes start executing jobs. The estimated repair time for the building's HVAC is 2 weeks.
  • T3-19 went down on Saturday night. Powercycling it on Sunday morning brought it back to life. However, syslog showed memory problems with the controller. The controller in T3-19 was replaced during Tuesday's fb1 downtime.
  • It turned out that the upper controller in 3510 (that was replaced by SUN a week ago) never actually came up. I would need another fb1 downtime to either try to reseat it or replace it again.
  • Got quotes for service contract for the old 15 ton unit from Star.

(Dwayne Giardina)

  • A few minor modifications to the badblocks monitoring script I wrote a week or so ago.
  • Labeling the new nodes.

Hanford

(Greg Mendell)

  • M7 Data was successfully archived at LHO. The times archived are: Start: 812746784 = Oct 07 2005 11:59:31 PDT = Oct 07 2005 18:59:31 UTC End: 812876416 = Oct 09 2005 00:00:03 PDT = Oct 09 2005 07:00:03 UTC Level 1, 3, and 4 RDS data sets were also successfully generated at LHO. There are no gaps in any of these data sets, and no errors occurred during archiving or reducing the data. A copy of this data has been or will be transferred to CIT, and then made available to Tier II centers via LDR. The new Level 4 RDS frames have 1024 s per frame file with separate frames for each IFO, and with these channels and downsampling factors for each IFO:
    • LSC-DARM_ERR 4
    • IFO-ACTIVITY_INDEX 1
    • IFO-ACTIVITY_STATE 1
    • IFO-ACTIVITY_TYPE 1
    • IFO-SV_STATE_VECTOR 1
  • The power and HVAC upgrade has finished at LHO, though we will be working to mitigate ambient noise from the new unit, and checking that it does not add any noise into the IFOs. It will take about another week to finish upgrading the cluster from 280 to 420 CPUs, as well as updating the operating systems.

(Ben Johnson)

  • Networked 16 of the new nodes. Currently performing some system imager tests (imaging node2, loading it on several nodes etc.).
  • Exported + Imported 142 tapes before M7. Strangely enough, I did not have to restart samd afterwards.
  • Worked with Duncan Brown on the segment database.
  • Fixed bad files at LLO, in /archive, that were caused by t3-9@LLO problems.
  • M7 went well this past weekend. There were no data gaps during the 36hour run.

General Computing (Wallace)

MIT

(Keith)

  • Ordered matlab licenses for several new users
  • Ordered replacement projector bulbs
  • Reinstalled thinkpad with latest o/s and patches
  • Tested external DVD burner with Fedora IV

LLO

(Dwayne)

  • Kantec software upgrade: had a few problems with the new version that took a couple days of trial and error testing possible solutions. Situation temporarily resolved.
  • Assisted Bonnie with more html documents.
  • Norton will not install on one laptop. Situation was reported this morning. I am looking at it today. Will advise when solved.

(Shannon)

  • Working on LDAP. I have all of the authentication encrypted, however the rest seems to go in the clear. Many of the Sun scripts for migrating data are set up for migrating from a generic vanilla LDAP installation, and not from a vanilla NIS+ conversion. Therefore, I am having to rewrite a lot of the scripts to massage the LDAP store into a compatible format. The current plan is to continue working on this and have it rolled out by the end of the week. At which point, I am sure bugs will pop up. Also, after it is deployed, I plan on documenting the entire process and distributing the notes to folks at MIT, CIT, and LHO so that they can possibly duplicate this process sometime in the future. This has been a long process.
  • Working with Sun to get a quote on a support contract for the messaging and directory server.
  • Found an infected machine late last night. After further inspection, it appears that the machine has been logged into an IRC channel and is being used as a DDoS bot. As far as I can tell, there has been no attempt to use this box as a jumping off point to infect other machines on site. I will have to investigate further via the snort logs to be sure of this. I will likely have Dwayne rebuild the machine after we learn the extent of the intrusion.

LHO

(Christine)

  • Trying to get problems with my Cisco contract sorted out so I can renew it. Then I can get help troubleshooting problems on the router with the new GigE connection. The backup network is at a standstill
    because we are unable to get a link between LHO and PNNL. With help from the PNNL engineer, I've tested each part of the circuit separately and it all tests OK, but when connected up we can't get a link. Amerion is throwing their engineer into the mix as of today, so we'll see what comes of that.
  • Helped a user with a printing problem which required reinstalling the network printer wizard and driver software.
  • Updating licensed software to the latest versions.
  • Set up accounts and networking for a long term visitors. Will have to setup a Sun workstation for his office.

CIT

(Veronica)

  • Project Science: Posted the presentations of the last workshop, other web updates.
  • CaJAGWR: User support.
  • LSC: Updates of the LSC publications database. Website updates.
  • LIGO: Website updates. Roster database updates. Installed security patches on Windows servers. On pictor, the attempt to run the update failed, I could not trace the error at the WinUpdate website. Following the advice at the msusenet.com I deleted the contents of %windir%\SoftwareDistribution, it seemed to solve it and I was able to install the patches. Worked with LHO on updates to the Advanced LIGO Schedule website, reset a couple of passwords, posted the current updates and gave Dwight an introduction on web publishing. Converted the footage of last week's NASA newsconference on short GRBs by Albert and others from analog to DV. Burned a couple of DVD copies.

(Mike)

  • Finished up loading Ansys 10.0 on engineering workstations.
  • Continued work on upgrading Norton Anti-Virus and Adobe 7.0 software for users.
  • Worked a problem with the VPN server. This turned out to be a DNS issue on the user's side of the house.
  • PICTOR: Is having some hardware issues that is causing this server to reboot it self. I am in the process of loading a new replacement server.
  • Finished up cleaning up a loaner laptop that came back from travel.
  • Spent a lot of time this week working on a parallel dongle issue that is used for Solid Works. I ended up downloading a driver from the Internet specifically for the dongle. The problem I was having is XP SR2 has a driver loaded into its Operating System for any HASP4 dongle. You have to disable that driver in order for the new driver to load. Aladdin website has a command line tool to disable the XP driver. I have ordered a USB Dongle to see how XP will handle going that route.
  • Working on getting old computers ready for surplus.
  • Other user misc. user support.

(Christian)

  • Barry Barish - Updated Barry's workstation to Norton 10 and Microsoft's latest patches.
  • 2flr W/B - Configured new workstation with the engineering Ligo image.
  • 3flr W/B - Updated two visitors workstation to Norton 10.
  • Veronica Kondrashov - Added 1GB of additional RAM to Veronica's workstation.
  • Meter - Updated two more workstation that I missed last week to Norton 10.
  • I went through 9 different loaner laptops imaging and updated them with the Standard Ligo image and Microsoft's latest patches.
  • Other misc.: Continued onsite software/phone support

(Larry)

  • Still working a number of procurement issues. The SUN contract is now through the system and we should have confirmation by the first of next week. Tracking down the dual-core systems. The companies have had problems getting a OS loaded on the units. So far the Enterprise versions of Linux, Windows XP64,  Solaris 10 and SUSE 10 appear to work on the machines. The units will ship with SUSE 10 and we will take it from there. Renewed a number of s/w maintenance contracts and still have a number to do.
  • Resolved a couple of end user issues. Setup a couple of new accounts and removed a couple of accounts.
  • Updated the backup s/w on one of the LDAS servers. Also, made a number of small changes on different severs to accommodate changes being made on LDAS machines.
  • Continued on a couple more of the complete backups on different servers. Still a few more to go.
  • Updated some of the DHCP tables and setups.
  • Assisted a number of users on updating their virus s/w and checking their systems for spyware.
  • Assisting the others and Ed C. clear up a number of equipment related items. Mostly, clearing out old items that people will not use anymore.
  • Finally, had a little time to do some work on documentation. The IP tables still have a lot of work needing to be done and I am in the process of updating the GC computer use policy.
  • Spent time tracking down e-mail issues. One issue concerning duplicate messages is still being worked on. The issue of missing messages has been resolved, the client utility had a trigger that was causing false positives for the end-user. Continue working on the mail-filters on the mail-server.

Mail Statistics for Oct 06 - Oct. 12, 2005

Mail Statistics

October 6 - 12, 2005

Rejected Messages

17,750

Virus Messages

979

False Positives

12

Accepted Messages

15,706

Total Messages

33,456

 

Advanced LIGO and Supporting R&D (Shoemaker)

Advanced LIGO Project Management

from Carol Wilkinson

Progress Updates

Progress updates for Advanced LIGO subsystem development for the period from September 1 through October 31 are due by November 10.

Subsystem leaders will receive update notices by October 24.

Meetings & Reviews

Future near term planned meetings & reviews are indicated in the table below.

The Ribbon/Fiber/Ear/Bonding PDR is in progress. Panel members have met to review documents and prepare questions for the SUS tem members.

Changes since last report are listed in blue.

Date

Sub sys.

Review

Topic(s)

Enabling event(s)

Schedule motivation

Status

Jul 11-13

SYS

SYS Mtg

CDS infrastructure & HAM Isolation Req’s

 

 

Report in progress

12-Jul
8-11 PT

SUS

PDR, Review 2

Electronics req & design; Focus is on the front end electronics (UK) -- limited Digital controls/electronics (US) review

 

 

Report in progress

~ Nov 4

SUS

SUS PDR

Review 8

Ribbon/Fiber/Ear/ Bonding PDR

Completion of the ribbon, fiber, and ear and fabrication and bonding design and processes

 

Panel first met Oct 12. Review scheduled for first week of Nov. Documents for review at

http://www.physics.gla.ac.uk/igr/sus/index.html

~Oct

SEI

HAM Critical Design Review

Recommendations w.r.t. HAM prototype development based on ETF results

Completion of SEI/BSC critical design reviews; LSC review of ASI HAM configuration design

Delay until requirements are revised. timely decision on proceeding with SEI/HAM prototype

 Delayed from Aug.

~Oct 12

SEI

HAM SAS  PDR

Review modeling, analysis, design for HAM soft system

Revised HAM SEI requirements.

Timely decision for proceeding with HAM SEI prototype.

 

~Oct

SYS

PDR, Review 1

Engineering & Implementation ('generic') Requirements;
Interfaces
Revised Optical Layout
Optomechanical Layout

completion of generic requirements definition; completion of first draft of ICD; revision to optical layout; establish integrated opto-mechanical equipment layout

timely system level definition enables/helps define subsystem reqmnts & design

 

~Dec 8

SUS

PDR, Review 3

Quad design

Completion of the quad controls prototype assembly; shipping to LASTI

timely transfer, to RAL & UB efforts, of lessons learned from the controls prototype

 Delayed until shipping to LASTI.

~Dec

SYS

PDR, Review 2

CDS Infrastructure
Stable Recycling Cavities
Lock Acquisition
Modulation Scheme
Power Induced Instability

Sufficient CDS requirements & concept work (also 7/11-13 mtg)
E2E Modeling for AL
40m Progress on Acq. & Mod.

CDS Infrastructure is key to subsystem electronics req.
Stable cavity is key to IO MMT design

 

~Jan

IO

PDR Review 1

PDD, Faraday Isolator, modulators, mode cleaners, mode matching design, Mach-Zender

SYS PDR?

 

 Work delayed on

FI, allows more

systems to be

reviewed

~Jan

SUS

PDR, Review 5

Triple design

Available SUS/US staff

Enable SUS/US final design phase

 Quad controls PT takes preced.

~Jan 25

AOS

AOS DRR/CD

Review 1

Stray light control, PO mirror and telescope, optical lever, initial alignment system

SYS PDR?

Delay stray light control review until other systems can be reviewed at the same time.

Delayed from Oct. to allow review of

add. systems.

~Feb

SUS

PDR, Review 4

Quad Installation

Completion of installation at LASTI

Inform the UK final design & noise prototype design effort ASAP

 Delayed shipping schedule

~Feb

SUS

PDR, Review 6

quad controls prototype test results
ribbon process/design

completion of LASTI testing

timely incorporation into final design effort on the noise prototype

 

~Feb

COC

PDR

Review 1

 Metrology, handling fixtures

 Review and select  vender and in-house metrology

Need handling fixtures for UK substrates arriving in Dec and Jan 05

Work delayed

~Mar

IO

PDR Review 2

Adaptive mode matching, Mode Matching Telescope

Determination of whether a stable recycling cavity will become part of the AL baseline; SYS PDR 2

 

 Waiting for stable

RC decision

TBD

AOS

AOS DRR/CD Review 2

 Thermal Comp., Photon drive

SYS PDR?

 

 Moved forward

 for timely finish

of AOS CDR’s

TBD

SUS

PDR, Review 7

BS, FM/ITM SUS design
RM design
non-cavity SUS

design work completion (has yet to start on FM/ITM, not mature for RM)

 

 

~May

COC

PDR

Review 2

 Cleaning Process, Optics prelim. design

SYS PDR?

 

 

Seismic Isolation

From: "Joseph A. Giaime" jgiaime@ligo.phys.lsu.edu

Agenda for the weekly SEI telecom

Friday, Oct 7, 2pm Eastern, 1pm central, 11am Pacific time

Announcements

  • Aaron Streets will be doing a rotation with the group at Stanford this quarter. Aaron is a first year grad student in Applied Physics, and will be working with us this quarter (i.e. this 'rotation') on some seismometer development questions.
  • New Mech. E. graduate student, Brett Shapiro, at MIT.  He will begin working on the quad suspension modeling and testing.

BSC SEI status, Ken M

  • First 3 bid packages received, comprising 80% of the structure.  The bid on the large part package was higher than expected, due to material cost changes; the others were as expected.  Delivery in early February.  Winner for small high volume parts is a new (to us) vender that specializes in 5 axis CNC work.
  • Ken is working to add holes to change the table to the 1" x 1" hole pattern; this will require a revised bid from the winner.  Extra cost will be due to more machine time and inserts.
  • Next set of bids, for the actuator mounts, locators and vacuum housings due in next Friday.
  • The revised blade and flex rod specs sent to the selected vender.
  • Ken has completed changing the drawings to accommodate the other ASI changes.  Richard F has been helpful, and has fixed some minor errors in the designs.

BSC work (Rich M)

  • 10 Hz amplification, Rubber Creep, and adaptive control

HAM control with VME (Pradeep)

  • Feedforward work is delayed due to the very noisy road construction work behind the lab; work only possible in late evenings.  Rich is working on a Simulink model to study the adaptive filter method.
  • Pradeep has the HAM HEPI horizontal loops working, but the verticals still not working.  SYS-ID is difficult due to the lab noise.

Electronics (Jay, Brian)  - the new readouts are in the GS13s, and installed in the ETF Tech Demo. The signals are all reasonable. Sys-ID to begin today, followed by damping.

  • Stanford has received the new GS-13 readouts.  Aaron and Matt have installed the boards in all of the feedback sensors, and the signals look acceptable.  Gains set so that ADC noise will start to limit at 3 or 4 hertz; whitening may be needed later.
  • Next: sys-id, redesign damping loops and control design.
  • Jay traveling.
  • Shyang awaits unlocker device from Stanford.

Other:

Norna notes that lower 3/4 of quad support structure ready, and upper part in the shop.  These will go to Stanford for the dynamic tests.

Suspension

From: Janeen Romie romie_j@ligo.caltech.edu

 

From: ctorrie <ctorrie@ligo.caltech.edu>

Mark and I now have a quad with reaction chain suspended in the lab. Still some work to do but both chains are stable!  See: -    http://131.215.114.242/view/index.shtml

From: Ken Mailand kmailand@ligo.caltech.edu

I will have a bearing test setup today for Calum for re.a quick test for the suspension fixture rotating table. The test using a piece of teflon between the plate contact surfaces requires too much break away force.

I’m working on a vertex BSC layout for Mike Smith that will show stay clear areas inside for the beam path.

From: Helena Armandula ahelena@ligo.caltech.edu

Adv. LIGO SUS - New air bake oven

Arranged for a JPL chemist to take FTIR samples from an aluminum plate that was cleaned and air baked following LIGO cleaning and baking procedures.  These tests will serve to qualify the oven.

In the past, the surfaces were sampled by wiping the surface with extracted fiber-free lens tissue using dichloromethane solvent.  This allows to determine the level and identity of molecular (oily) contamination on the surface.

From: Jay Heefner <jay@ligo.caltech.edu>

AdL SUS

The preliminary design of the AI chassis needed for the quad controls prototype is complete. The AA chassis is in the works.

Initial measurements of the noise of the PCI-X ADC module shows an input referred noise of ~3uV/rtHz with a 40Vp-p dynamic range.  DAC measurements are in progress.

Discussions last week with the UK electronics group went very well. Over the next several weeks we will try to finalize the electronics requirements and the motivation for the requirements. This will include an estimate for the required whitening and dewhitening.

From: Rolf Bork <rolf@ligo.caltech.edu>

  • Got PCI-X ADC and DAC drivers working for the quad controls.
  • Quad controller data acq is working via the VMIC5579 reflected memory, which is compatible with what is presently installed at Lasti.

Core Optics

From: Bill Kells <kells@ligo.caltech.edu>

I am now gearing the PI work toward suammarizing the situation for the NSF review. Most recently a new twist: it had been unresolved why my model and the BSV (BRag, STregin, Vychanin). model of PI were in considerable discrepancy.  Finally I have resolved this. There is NO discrepancy! What was wrong was the Perth groups use (apparently in all their papers ....??) of the BSV analysis. Since The Perth people were here this summer I was strongly influenced by their adaptation and, in fact, was making comparision that rather than the BSV original. So its good now that there is no puzzle.

I have now been focusing on what exactly would we expect if there is a PI threshold exceeded in a real IFO.

Pre-Stabilized Laser

From: Peter King pking@ligo.caltech.edu

An hour meter circuit was conjured up for the Innolight NPRO.

University of Adelaide

  • Installed a 10-W laser in TAMA, this replaces the laser purchased from Sony.  The laser is an air-cooled slab design.
  • Work continues on a 100-W laser.  At the moment they obtain about 50 W from a few hundred watts of pump power.  However there are some slab mounting issues.  The end pump slab absorbs approximately 90% of the pump power.  However there is a need to optimize the thermal lensing and cooling.  The geometry employs a zig-zag path to increase the absorption length.

Laser Zentrum Hannover

  • The 150-W laser is back up and running, although there are some problems with the power scaling.  It is believed that the problems encountered with the laser at the moment are due to some damage on the rods.  Some spots on the anti-reflection coatings are noticeable.  Experiments with the position of the quartz rotator showed that as the quartz rotator was adjusted to improve the alignment into the high power stage, the polarization changed from being linear to elliptical.  However more output power is obtained, although the polarization is unknown.  The quartz polarizer is no longer adjusted to improve the alignment.

Stanford

  • The two end-pumped slabs are up and running.  The alignment of the system is being tweaked to improve the performance.  Pump diodes from Limo and LaserLine are being evaluated.

AEI

  • The latest relative intensity noise (RIN) measurement result is a RIN of 4.0E-9 at 200 Hz.  Careful studies were made of the photodetectors used, servo electronics and any scattered light.  Thus far the limitation appears to be associated with the semiconductor the photodiode is fabricated from.  The results obtained with InGaAs are better than those obtained with Si photodiodes.

Auxiliary Optics

From: Michael Smith <smith@ligo.caltech.edu>

SLC

Made a preliminary layout of a ring heater using Solid Works in accordance with Phil Willems calculations of 10/6/05. The ring heater design fits within the "stay-clear" area of the ITM SUS.

For additional information about this report, contact Stan Whitcomb or Phil Lindquist