From: Stephen Sanders (ssanders@ku.edu)
Date: Wed Jun 25 2003 - 19:56:40 EDT
Hi, I just submitted a problem ticket to rcf for what appears to be a problem with /brahms/data07. I ...steve RCF Call Tracking System New View Followup Search Help View problem ticket 8851, Original Submission View Full Ticket Followup Problem 8851, Original Ticket - Open , Priority: Medium Customer: Stephen Sanders , x7681, High Energy And Nuclear Physics Directorate Submitted On: Wed, 25 Jun 2003 19:49:02 Submitted By: Stephen Sanders , x7681 Problem Type: File Services Owner: Maurice Askinazi , x2159 Summary: disk file corruption Collaboration=Brahms I have been creating GEANT output files for subsequent "replay" on /brahms/data07. The files are created with LSF batch jobs. These files seem to be getting corrupted with subsequent disk activity. As a test I created a short file /brahms/data07/sanders/simu/gbrahms_output/sim_187.cdat. I tested this file using a simple program to dump the geant file header and the output looked fine. I also made a duplicate of the file, which also tested fine. After starting some additional geant simulations, I went back to test the original files and found that some of the header information seemed to have been scrambled. The duplicate file also seemed to be corrupted. The size of the files did not change. The short run was done after encountering problems over the past day of programs that seems to be working suddenly failing. I might also note that older files that I created on the disk seems to continue to work fine.
This archive was generated by hypermail 2.1.5 : Wed Jun 25 2003 - 19:58:00 EDT