Minutes for the terameeting of October 15 2013 at 14:30

Present: Patrick T., Olivier, Nawfel, Henry

Action: Nawfel / Patrick will examine the source code for the generation of WCS

Action for Henry: Need to contact Wei-Hao and friends for the Centers of the stacks

Action for Henry: double check that the limit is still in place and do a test download

Next meeting 31.10.13 at 9:30

Status S82 (Nawfel)

 Images which had the grade C were rejected. . Also some 18 images B were rejected. Around 10% of images were rejected (1100 images in total).

 Still problem with qualityFITS. WCS of images is sometimes wrong in the weight maps. The header is not created correctly for the weight map for certain images. Don't know why.

 Should take one image and see if can fix this problem. This is a big problem as it can produce weight maps which are very large

 Patrick already changed from fits2head to dfits to avoid this problem, but it seems that this problem already existed. Need to change the method that the weight map header is generated from the input image.

 Steps remaining: qualityFITS / scamp / swarp.

PI processing

 Nawfel has almost finished processing WIRCAM PI data 214 images maybe
 Pagani data: need to see Pagani: to define final data products

CFHTLS deep re-processing
 Valerie de L. would like to have a new deep stacks with better sky subtraction. On-going task.

Status VIKING (Patrick)
 It seems that the request limit is back at 1000 files, so we cannot download the other bands in CFHTLS
 Henry mentions that other people have noticed problem with VIKING + CFHTLS

Status Ultravista DR2 (Henry)
 Henry is proceeding with messenger article which will be published in December

Status TERAPIX/COSMIX cluster (Patrick)

 Patrick has made backups of COSMIX1,2,3,4 on fcix4
 Also made backups of fcix2,mix7,8,9,10 on xxx
 These machines can be stopped any time.
 NGVS machine will not be accessible to TERAPIX users if
 only machines will be cosmix5 and cosmix
 1u cosmix
 Should we think about turning off the other cosmix machines?
 mix31 (zeemax machines) can be added back to TERAPIX cluster

Status PUMAS (Fred/Patrick)

 Limit of end of October is IAP limit real limit is end of October
 Should be able to stretch things for a few more days, waiting for an announcement at the end of the year.
 Machines have to be delivered before the end of October.

Status TERAPIX2 cluster (Patrick)

 Nothing new for the new servers / SSD installed
 Still no news about the new accounts on the machines: if nothing today then will send a mail.
 Patrick asked if IAPIX could be made visible from machines other than IAP.
 No condor / no ganglia / home but no cross-mounting / mix33 (high memory machine) still crashes (error code cryptic). Could be a problem of power supply.
 Cables are not super-stable state
 Mini-KVM in new machines allows remote administration

Status GTO Bo (Olivier)
 All data processed action closed