From: Boudewijn Ambrosius Subject: [IGSMAIL-1217] KOSG data available again Date: Thu, 15 Feb 1996 4:47:28 PST ****************************************************************************** IGS Electronic Mail Thu Feb 15 4:47:28 PST 1996 Message Number 1217 ****************************************************************************** Author: Boudewijn Ambrosius Subject: KOSG data available again Station KOSG is back on line. Since the forced receiver change from an SNR-8 to an SNR-12 on January 26, the operations and data handling have been hampered by several unforeseen problems. When the change took place, only a 1Mb flashcard was available and it was realized not until after a few days that this would just hold about half a day of data. This problem was obscured by another problem with our RINEXing software that could not handle more than 10 receiver channels. After these problems were corrected by downloading 4 files per day and installing an updated version of the RINEXing software, we distributed the data of day 34 - 39 on February 9th. After this we found that the RINEX files of days 35 and 36 were smaller than the others, although the downloaded Conan Binary (CB) files were about the same size. This was traced to occasional failures of the CB decompression software to fully decompress some of the 6-hourly files. First the decompression software was suspected, but we now know that the problem is in some of the downloaded CB files. It is not yet clear what exactly causes these files to be corrupted occasionally. This seemed to be a random problem, until the automatically downloaded receiver files gradually started to become smaller every day during the last 3 days. They seem to be truncated. When we downloaded the last file, which originally was about half the size it should be, a second time under manual control it was OK. All we can do for the moment is to monitor the problem and to swap hardware if the problems persist. Because of all these problems, we released the data for days 40 - 44 only yesterday, February 14th. The data of days 34 -39 have also been resubmitted today after reprocessing with the latest versions of the CB decompression software and the RINEXing software. On the one hand this may have resulted in less data because the decompression software now stops after the first erroneous byte in the CB files (as opposed to the older version which would sometimes continue and find good data after the trouble spot), but on the other hand there may be more data due to the larger number of channels allowed by the new RINEXing software (we've seen as many as 11 satellites being tracked simultaneously!!). Furthermore, we have now also submitted the data for days 26 - 33. These are only partial files (2nd half of the day), because there was only one download per day during this period, which was too slow with respect to the limited capacity of the 1MB flashcard. Finally, the Rubidium frequency standard has been reconnected to the SNR-12, around 12:30 GMT yesterday, February 14th. We hope the situation has now stabilized. Boudewijn Ambrosius [Mailed From: Boudewijn Ambrosius ]