Re: CRS instabilities??

From: Stephen J. Sanders (ssanders@ku.edu)
Date: Wed Apr 17 2002 - 22:41:34 EDT

  • Next message: Stephen J. Sanders: "Re: CRS instabilities??"

    OK, I'll report another strangeness:
    
    I've been trying to replay the tofw data to look at the correlation of the
    number of hits with  centrality.  I've developed a module that is working
    fine on my local ppc (macintosh).  On rcas the geo manager seems to be
    screwed up: I'm getting multiple errors of the type
    
    TOFW slat number 22 is out of bound
    TOFW slat number 23 is out of bound
    TOFW slat number 24 is out of bound
    TOFW slat number 25 is out of bound
    TOFW slat number 26 is out of bound
    ...etc
    
    
    Also, if I try to access the geo DB for TPM1 I get:
    Warning in <BrGeometryDbManager::GetDetectorVolume>: For TPM1 Failed
    
    On my local system I'm not seeing this problem.
    I stopped using the local .geo ascii files on my home system since it looked like
    the DB was working.  Thinking that I might have managed to get different code on rcas from 
    what I'm using locally, I deleted my brat and personal module directory as well as my bratmain script on
    rcas and did an scp transfer from home to rcas, then recompiled.  I'm at a loss as to
    what I might have screwed up. My next step will be to go back to the ascii geometry files, but
    I haven't checked this yet.
    
    ...steve
    
    
    Ian Bearden wrote:
    
    >Hello Dev-iants,
    >I have been trying (and trying..) to (re-re-re-re-re-)reduce data.  I am not
    >having such a good time of it.  For the past week or so, my guess is since
    >the upgrade to RH7.2, I have been losing 10-30% of the sequences due to the
    >reduction code crashing.
    >I am now running with much higher verbosity and debug levels, which I hope
    >will help.  While this is ongoing, though, I though I would just see if
    >anyone else has had similar problems in the past few days.  I had hoped that
    >the problems would disappear with new BRAT version, but they do not.  Could
    >this be because of the problems with gcc2.96?  I don't have any really good
    >ideas, but surely some of ya'll know more than I do...
    >Ian
    >
    



    This archive was generated by hypermail 2b30 : Wed Apr 17 2002 - 22:42:08 EDT