Teraminutes for the meeting of August 27th 2004

Present: everyone except Gilles and Gerard

dotProject (all:)

 Everyone is encouraged to fill out the dotproject tasks and to indicate their vacation periods and pending tasks.

Spica (Laurent).

 Laurent has added more information in the survey processing section of Spica, which sorts the amount of processing pending by number of images in each stack. Now for each stacked image a sub-section of the stacked image is available full res. as a PNG and rice-compressed fits image.

 Other work has been carried out to monitor the total amount of disk space required for Spica processing.

 Henry will compare the photometric properties of a stacked image from Spica with one which was independently reduced outside the pipeline.

 Photometric scaling: some thought required how photometric image is selected. Emmanuel will provide an output plot from scamp showing the variation of zero-point as a function of time. For the images which there is no photometric redshift we take the median PHOT_C before/after the images. Henry suggests to select the photometric 'reference' image as the one which has least extinction.

QualityFits (Fred): PNG bitmaps are not generated for some images. Don't yet know why this happens. Fred will install a new release of the ImageMagic tool.

Astrometrix (Laurent, Fred): Fred will co-ordinate with Laurent to provide the same version of astrometrix for all machines in the cluster (both Opterons and athlons)

Scamp (Emmanuel) What's been done: bug corrected in the management of the variation in degrees of freedom in Scamp fixed.

 No solution yet found for the blockages found by Yannick: in these cases, Scamp uses CPU but no progress is made in computing the astrometric solution.

 Several other tasks still needs completed, for example to carry out correct relative weighting between reference catalogue and internal sources.

Panorapix (Fred): Fred will work ONLY on Panorapix on Wednesday, in attempt to make some progress before the next astro-wise telecon. Any requests for work on the cluster will be deferred until then.

Skywatcher (JC) People can send JC suggestions, requests for functionality to JC. A test version is available at http://malaprix.iap.fr:8080/skywatcher .

 Still problems with CSS and optimization with WCS. Optimize refreshing the image by carrying out drawing tasks in the background.

Poster (Yannick, Emmanuel) Logos will be overlaid on the image before to be sending it to the steering group for authorization (CFHTLS - TERAPIX
 CNRS)

PI data reductions (Yannick, Mireille): For the Dougados data there are 378 images in class A-B. Zones are to be identified and images to stacked. Laurent will make a special version of Spica to handle this process.

 Mireille has processed all 03BF12 images (V. Cayatte data) and generated stacked images and catalogues. Last checks will be done over the WE and data will be released to Veronique on Monday. Still problem with the u-band calibration.

Hardware (Fred, Yannick, Emmanuel):

 The fibre-channel array has been ordered and should arrive early next week. T

 This week, machines in the cluster were re-arranged to provide more space for the new array and for the new quad-opteron (mix3). mix3 is very stable: no crashes to date!

Remaining cash for hardware: There are 730eur in AVO funds which must be spent before the end of October.

 Around 15,000 eur left which will be spent on buying extra disks for mix3, which we will try to delay as much as possible, at least until the end of the year.

Heat load in the computer room: Apparently the computer room is already at the thermal limit: TERAPIX machines alone consume 8kw of electricity out of 32kw total.

 The room is rated at 32 kw maximum capacity. Question from fred: does setiathome raise the overall temperature of the room? Fred proposes terminating all setiathome tasks and observing if there is a correlation with the room temperature.

 Yannick is concerned about safety in the computer room, and that plugs are not overloaded.

 We will try to remove ev6 as soon as possible to reduce the heat load.

 Fred and Gilles will install automatic shutdown of Terapix hardware when the room temperature exceeds the limit (30 deg suggested).