home *** CD-ROM | disk | FTP | other *** search
/ NetNews Usenet Archive 1992 #30 / NN_1992_30.iso / spool / comp / os / vms / 19506 < prev    next >
Encoding:
Text File  |  1992-12-20  |  3.5 KB  |  100 lines
  1. Newsgroups: comp.os.vms
  2. Path: sparky!uunet!zaphod.mps.ohio-state.edu!saimiri.primate.wisc.edu!zazen!psl.wisc.edu!chsra2.chsra.wisc.edu!DSPENCER
  3. From: dspencer@chsra2.chsra.wisc.edu
  4. Subject: $convert doesn't pack this file
  5. Message-ID: <1992Dec17.173404.9461@pslu1.psl.wisc.edu>
  6. Sender: news@pslu1.psl.wisc.edu (USENET News System)
  7. Reply-To: spencer@macc.wisc.edu
  8. Organization: CHSRA/UW-MADISON, MADISON, WI
  9. Date: Thu, 17 Dec 92 17:34:04 GMT
  10. Lines: 88
  11.  
  12. I have been trying to reduce the size of a large indexed file (FDL included below).  I run $anal/rms/fdl, and generate the listing on the left side of the post, with a DATA_FILL factor of 57% in the anal_of_key 0 section.  I then run
  13. $edit/fdl/anal= /script=opt  
  14. on the *.fdl file, and it generates the three new allocations I have inserted
  15. on the right.  Then I run $convert..., but the file that is generated using 
  16. the new .fdl file is the same size as the old one.  I thought $convert was 
  17. supposed to tune and reduce files in this situation.  Am I wrong?  Any
  18. suggestions for improving the DATA_FILL I get out of this situation?
  19.  
  20. SYSTEM
  21.         SOURCE                  VAX/VMS
  22.  
  23. FILE
  24.         ALLOCATION              864168
  25.         BEST_TRY_CONTIGUOUS     yes
  26.         BUCKET_SIZE             8
  27.         CLUSTER_SIZE            8
  28.         CONTIGUOUS              no
  29.         EXTENSION               65535
  30.         FILE_MONITORING         no
  31.         GLOBAL_BUFFER_COUNT     0
  32.         NAME
  33.         ORGANIZATION            indexed
  34.         OWNER
  35.         PROTECTION
  36.  
  37. RECORD
  38.         BLOCK_SPAN              yes
  39.         CARRIAGE_CONTROL        carriage_return
  40.         FORMAT                  fixed
  41.         SIZE                    37
  42.                                                    ****edit/fdl/optimized
  43. AREA 0                                              changes this to
  44.         ALLOCATION              862488                  475392
  45.         BEST_TRY_CONTIGUOUS     yes
  46.         BUCKET_SIZE             8
  47.         EXTENSION               65535
  48.  
  49. AREA 1
  50.         ALLOCATION              1680                     1568
  51.         BEST_TRY_CONTIGUOUS     yes
  52.         BUCKET_SIZE             8
  53.         EXTENSION               424                       392
  54.  
  55. KEY 0
  56.         CHANGES                 no
  57.         DATA_KEY_COMPRESSION    yes
  58.         DATA_RECORD_COMPRESSION no
  59.         DATA_AREA               0
  60.         DATA_FILL               100
  61.         DUPLICATES              yes
  62.         INDEX_AREA              1
  63.         INDEX_COMPRESSION       yes
  64.         INDEX_FILL              100
  65.         LEVEL1_INDEX_AREA       1
  66.         NAME                    "RID"
  67.         NULL_KEY                no
  68.         PROLOG                  3
  69.               SEG0_LENGTH             10
  70.         SEG0_POSITION           0
  71.         TYPE                    string
  72.  
  73. ANALYSIS_OF_AREA 0
  74.         RECLAIMED_SPACE         0
  75.  
  76. ANALYSIS_OF_AREA 1
  77.         RECLAIMED_SPACE         0
  78.  
  79. ANALYSIS_OF_KEY 0
  80.         DATA_FILL               57
  81.         DATA_KEY_COMPRESSION    73
  82.         DATA_RECORD_COMPRESSION 0
  83.         DATA_RECORD_COUNT       5882786
  84.         DATA_SPACE_OCCUPIED     820736
  85.         DEPTH                   2
  86.         INDEX_COMPRESSION       28
  87.         INDEX_FILL              98
  88.         INDEX_SPACE_OCCUPIED    1040
  89.         LEVEL1_RECORD_COUNT     56737
  90.         MEAN_DATA_LENGTH        37
  91.         MEAN_INDEX_LENGTH       13
  92.  
  93.  
  94. Please use the address below if you email, the address in the header will
  95. not be correct until early next year.
  96. -----------------------------------------------
  97. Dan Spencer         spencer@vms.macc.wisc.edu
  98. Center for Health Systems Research and Analysis
  99. UW-Madison
  100.