
[Science] [BAS home] [Met home] [Beowulf home]	Antarctic Meteorology

Mon Feb 10 17:01:53 GMT 2003 | jobsheet

yabga ([cp yabfe] Control ; 4.5)

Start time: 1978/2/1, length: 30/0/5
Processors: 1 * 1, Exec name: $RUNID.exe; Timestep (mins): 30
Start dumps: $DUMPS/Dec01_1994.clim
Fortran mods
Diags
Userprestash files
Status: [good]
See-also:
Notes:
1. Notes go here...
History: yabga -> yabfe -> yabce -> d -> c -> yabba -> yaaug -> e -> d -> yaapb -> a -> [Upload std xaaaa]

Purpose of Run (long)

This is a control run, not really ever run much, except to do bit comparison tests.

Results

Bit comparison achieved. Details to follow (ha ha).

32 bit

OK. It compares between 1x1 and 1x2, *with* my qtpos1a fix:

master:~/PUM_Output/vn4.5/dataw.yabga$ cumf.ksh 1x2/yabgaa.dah82b0 1x1/yabgaa.dah82b0
Summary in:                       ,/home/wmc_32/tmp/cumf_summ.13276
Full output in                    ,/home/wmc_32/tmp/cumf_full.13276
Difference maps (if available) in:,/home/wmc_32/tmp/cumf_diff.13276
  COMPARE - SUMMARY MODE
 -----------------------
  
Number of fields in file 1 =   290
Number of fields in file 2 =   290
Number of fields compared  =   290
  
 FIXED LENGTH HEADER:        Number of differences =  4
 INTEGER HEADER:             Number of differences =  0
 REAL HEADER:                Number of differences =  0
 LEVEL DEPENDENT CONSTANTS:  Number of differences =  0
 ROW DEPENDENT CONSTANTS:    Number of differences =  0
 LOOKUP:                     Number of differences =  0
 DATA FIELDS:                Number of fields with differences =  0
  files compare (ignoring Fixed Length Header)

I've also checked out to 10 days and it compares OK *without* my qtpos1a fix, and without gcom2.9b5.

64 bit

With gcom2.9b5, it is OK (at least up to 4 proc). With the original gcom, without my qtpos1a fix, it isn't: it fails bit-cf early on, certainly on the first days average. With the original gcom, or gcom2.8, *with* my qtpos1a fix, its OK.

Here is the failure with original gcom, and no fix:

master:~/PUM_Output/vn4.5/dataw.yabga$ cumf.ksh yabgaa.pah8feb 1x2/yabgaa.pah8feb

Summary in:                       ,/home/wmc/tmp/cumf_summ.14501
Full output in                    ,/home/wmc/tmp/cumf_full.14501
Difference maps (if available) in:,/home/wmc/tmp/cumf_diff.14501
  COMPARE - SUMMARY MODE
 -----------------------
  
Number of fields in file 1 =     1
Number of fields in file 2 =     1
Number of fields compared  =     1
  
 FIXED LENGTH HEADER:        Number of differences =  0
 INTEGER HEADER:             Number of differences =  0
 REAL HEADER:                Number of differences =  0
 LEVEL DEPENDENT CONSTANTS:  Number of differences =  0
 LOOKUP:                     Number of differences =  0
 DATA FIELDS:                Number of fields with differences =  1

Field     1 : Stash Code    24 : SURFACE TEMPERATURE AFTER TIMESTEP   : Number of differences =     3150
  files DO NOT compare

Past last modified: 21/3/2003 / wmc@bas.ac.uk