Minutes for the Terameeting of January 12th, 2009!. Present: Yannick, Henry, Yuliana, Fred, Mathias (M.), Mathais (S. - phone), Patrick, Greg.

Yannick's summary document

Welcome Yuliana

This is the first meeting since Yuliana joined TERAPIX. Yuliana is using 08AC24 MegaCam data as a training set.

 Yuliana had problems running Condor on the cluster and problems with mix17 which crashed.
 Patrick and Yuliana have added ways to detect the saturated pixels in the images; it seems until this point saturated pixels were not correctly detected in the images. QualityFITS has been modified and after testing the modified version will be installed on the cluster.

Once this training phase is over then Yuliana will work more directly with Mathias and Greg in using Youpi to process data on the cluster

Megacam PI: Mathias, Yannick

 Yannick will verify that the Youpi QF tables are identical to those produced manually
 Masks have been converted into full WCS format (in the past the borders were in pixel co-ordinates). All masks are now completely in WCS format.
 T0006 data on cluster
 Yannick asks if Mathias (S.) could investigate the feasibility of using R to compute the field-to-field offsets based on the positions of the stellar locii.

WIRCAM PI (Patrick, Henry)

 Hudson responded and he was happy to get all data reduced before Christmas.
 Flaguet (pillars of creation). Unfortunately he didn't observe any 'off' field so the sky subtraction will be problematic.
 It was necessary to make some changes to sextractor so that images with many pixels at zero could be correctly processed (a large fraction the sky is flaed. This is for Bouvier (Serpens) data.
 134 images of Mohaddesseh Azimlu have been downloaded.
 WIRDS/WUDS: Henry asks which is the more important. WIRDS is top priority at the moment.

Youpi (Mathias, Greg)

 Small demo from Mathias on latest improvements to Youpi, including thumbnail generation using new library.

Development (Background)
 Patrick modified sky-subtraction task [details?]

Hardware (Fred)

 mix17, mix18 processors have been changed.
 mix15 has a problem with the motherboard, only 4gb of memory is
 working. The motherboard / memory modules will have to be returned.
 New machines have been installed in the racks and powered on, but no OS has been loaded yet. Fred would like to make some tests with a new version of linux
 Hardware attack: fcix3 compromised because of a week password in the user ID 'idris'; this user ID has now been removed. (idris/idris). Need to reinstall system, maybe use scientific linux (which is used in Fermilab and CERN).
 Problem with one KVM (port not working). *for the moment the cluster nodes are not accessible by KVM*
 fcix3 is available but only within the cluster (outside accesses are disabled).
 mix3 motherboard has returned, still to be installed.

Other matters

 The VISTA telescope and the UltraVISTA project have encountered significant delays. It seems unlikely that there will be data before the end of 2009.
 33 keur from INSU awarded (+ 20 keur from PNC/ANR proposals which has yet to be decided. We will find out about these funding decisions in June).
 This year is the last year we will be funded for the CFHTLS, but Yannick got feedback from INSU/CSAA that terapix will still be funded as a National Data Center.