From Habrich at t-online.de Wed Nov 30 03:41:08 2005 From: Habrich at t-online.de (Habrich at t-online.de) Date: Wed, 30 Nov 2005 12:41:08 +0100 Subject: [IGSMAIL-5266]: Concatenation of hourly files at BKG data center Message-ID: <1EhQKi-1TLXIO0@fwd32.aul.t-online.de> ****************************************************************************** IGS Electronic Mail 30 Nov 03:41:35 PST 2005 Message Number 5266 ****************************************************************************** Author: Heinz Habrich This e-mail is an information for IGS-GDCs, IGS-ACs and all other users of the BKG-RDC about the strategy to concatenate hourly files that is applied at BKG. Since the launch of the new data center server on July 27, 2005 we apply the following "rule" for concatenation of hourly files: Each time when an hourly file has been uploaded to BKG, we just add it on the already existing concatenated file. There were two reasons for us to establish this procedure: 1) The handling of incoming hourly files is very clear and could easily be explained to outstanding people. Every time when an hourly file has been uploaded to BKG, we need just to add it on the already existing concatenated file. 2) We want to provide all incoming observations as soon as possible to the users. We have therefore no longer a "scheduled" processing of incoming files, but process every incoming file as soon as the upload has been completed. If you concatenate the hourly files only once a day, you have to check, if all 24 files has been arrived and in the case that one file is missing you have to wait (according to a time period for waiting that has to be defined before). At BKG we avoid any kind of "waiting" whenever possible. This strategy has some influences on the available daily data files, for the period shortly after 0:00 UTC. The following example will explain the situation: Between 00:00 UTC and 01:00 UTC of day N there are typically 3 versions of the daily observation file available: 1) between 00:00 and approx. 00:04 : An obs-file that was concatenated from hourly files and contains the observations from 00:00 to 23:00 of the previous day (day N-1). -> status of that file: Last hour is missing! 2) between approx. 00:04 and ,e.g., 00:46 : An obs-file that was concatenated from 24 hourly files and contains all observations from the previous day. -> status of that file: OK ! 3) after approx. 00:46: A daily file that was directly sent by the operational center and has replaced the daily file, which was available before. -> status of that file: OK ! The following list of files identifies all affected stations: bogi3340.05d.Z hers3340.05d.Z mate3340.05d.Z reyk3340.05d.Z unfe3340.05d.Z bor13340.05d.Z hert3340.05d.Z medi3340.05d.Z reyz3340.05d.Z unsa3340.05d.Z brst3340.05d.Z hflk3340.05d.Z mets3340.05d.Z riog3340.05d.Z vene3340.05d.Z brus3340.05d.Z hofn3340.05d.Z mizu3340.05d.Z sass3340.05d.Z vis03340.05d.Z bucu3340.05d.Z hueg3340.05d.Z morp3340.05d.Z scor3340.05d.Z warn3340.05d.Z bzrg3340.05d.Z ieng3340.05d.Z nico3340.05d.Z scub3340.05d.Z wroc3340.05d.Z cagl3340.05d.Z ista3340.05d.Z not13340.05d.Z sfer3340.05d.Z wtzj3340.05d.Z cagz3340.05d.Z joz23340.05d.Z obe23340.05d.Z sofi3340.05d.Z wtzr3340.05d.Z conz3340.05d.Z joze3340.05d.Z onsa3340.05d.Z spt03340.05d.Z wtzz3340.05d.Z dlft3340.05d.Z khar3340.05d.Z orid3340.05d.Z sulp3340.05d.Z yebe3340.05d.Z drej3340.05d.Z kir03340.05d.Z ous23340.05d.Z sutm3340.05d.Z zimj3340.05d.Z dubr3340.05d.Z lama3340.05d.Z pado3340.05d.Z tash3340.05d.Z zimm3340.05d.Z ffmj3340.05d.Z leij3340.05d.Z pdel3340.05d.Z thu23340.05d.Z zwe23340.05d.Z geno3340.05d.Z lpgs3340.05d.Z penc3340.05d.Z thu33340.05d.Z glsv3340.05d.Z mar63340.05d.Z pots3340.05d.Z titz3340.05d.Z gope3340.05d.Z mars3340.05d.Z qaq13340.05d.Z tlse3340.05d.Z gras3340.05d.Z mat13340.05d.Z rabt3340.05d.Z ulab3340.05d.Z We ask all users to check the availability of daily files after 00:05 for download. The upload from BKG to GDCs (e.g., to CDDIS) will automatically be done as soon as the daily file has been replaced at BKG.