Minutes of terameeting of the 26th May 2009. Present: Fred, Yuliana, Patrick, Greg, Yannick, Henry. By remote: Mathias M. and Mathias S.

Summary document circulated by Yannick before the meeting

Status of Youpi: Mathias M., Mathias S., Yannick, Yuliana
 Permissions have been implemented in Youpi and now several users can share processing and data products.
 Greg and Mathias M. are investigating the automatic generation of configuration files in order to process the wide frames.
 Mathias M. thinks we are ready to start production unfortunately the transfer of data from mix21 to fcix5 is very slow.
 Unfortunately the cache is off in fcix5 is off again. It could be a problem with the battery unit. This is a why the transfer the is slow -- need to find out what the problem is .
 300gb in five hours copied, not a lot: it's very slow.
 After ingestion is finished we need to stop fcix5 because there is still a problem with a cache: could it be a problem with battery backup with the cache.
 Preingestion is done for the wide and has worked.
 scripting and real-time checksum md5 on CADC data can be done by greg.
 For the minute checksums computed by the pipeline are actually different; need to fix this.
 Two validated images are missing at CADC
 Total amount of data around 3.9Tb of input files.
 W4+W2 are now ingested without problems.
 W3 has started. Could be finished in the next four hours.
 There are 686 images which have been QFITSed In W4 - greg will create a list with
 Yuliana will run QFITS
 Next step will be the creation of .ahead files (which will happen outside .
 Then we will need to run the new scamp.
 For the SWARP all the input lists exists and the config fies will be generated automatically.
 Remaining steps in processing reviewed by Yannick.
 Data transfer to CADC will be with uncompressed data.
 Yannick hopes to complete the whole process by June 15th.
 Yannick has discussed the schedule between 25th of May until September; Mathias M. this thinks this possible.
 deadline for processing by wide by Aug. 15th
 Deep by Sept. 15th.

[Hardware (Fred)
 New mixes are configured but QF still not running correctly on them (new version of QF needs to be installed on all nodes)
 Currently problems with mix19 and fcix5
 The mix19 crashes seem to be related with the 3ware card. We will order a new raid card as soon as possible (500 euros).
 64 gig should be enough for all the wide fields
 128 gig / 256 gig machine available at planck
 qfits not running on the new machines correctly yet, need to fix this.
 also Patrick has problems with his scripts.

WIRCAM (Patrick and Henry)
 Henry will to reduce all the COSMOS-H data for the meeting in June: Deadline, 5th of June.
 10,000 wircam images have to be resample 0.15''/pixel images.
 Rich Bielby is processing all the D3 JHK images.
 Photozeds better now in the D1 field after inclusion of data from T03 and removal of the data from T05 (rectangular region is no longer problematic).
 WUDS YJ sky subtraction finished, Ks is next up.
 Dubus H reduced and stacked.
 Patrick mentions that "Strange PI request" data no feedback from PI.

Youi (greg)
 Several bug fixes / improvements from Greg

Future purchases
 48,000 (CNRS) + 9,000 (NGVS)
 Reserved funds 1000 (Yuliana, parcels) , 500 euros raid mix19
 Buy two desktops (Grenet + Yannick?)
 Need to buy a new fcix6, 5 new nodes.
 20,000 WUDS for next year.
 14,000 ECOSSTAT
 There may be around 10,000 for ANR-DESIR will be set aside for next year.

Other matters
 Next terameeting next week.
 Henry will investigate STIFF parameters for T05 for preparation of posters for RER-B

Todo

 Buy raid card for mix19
 Buy replacement motherboard for Barney Rowe's desktop.