[MITgcm-devel] mnc and thsice

[Daniel Enderton]

Hey Ed,

There is a problem with the recent mnc-ifying of the thsice package. 
When the coupled system (verification/cpl_aim+ocn) is run without 
MNC, everything runs fine, but when run with MNC, there is a crash 
when writing the time averaged thsice data.  The STDERR files are 
zero size but the following errors appear:

[enderton at ao Fail1]$ tail -100 C28_Eq_Ex_Edge.e7148 | more

forrtl: severe (66): output statement overflows record, unit -5, file 
Internal Formatted Write
Image              PC        Routine            Line        Source            
mitgcmuv.A6        0823D33C  Unknown               Unknown  Unknown

Stack trace terminated abnormally.


[enderton at ao Fail1]$ tail -100 C28_Eq_Ex_Edge.o7148 | more

p1_2967:  p4_error: net_recv read:  probable EOF on socket: 1
rm_l_2_15140:  p4_error: listener select: -1
P4 procgroup file is pr_group.
rm_l_3_15159:  p4_error: listener select: -1
rm_l_4_23014:  p4_error: listener select: -1
rm_l_5_23033:  p4_error: listener select: -1
rm_l_6_14834:  p4_error: listener select: -1
rm_l_7_14853:  p4_error: listener select: -1

[enderton at ao STD]$ tail -100 STDOUT.0000000080.0002 | more
...
BBerr - Bug: IceT(2) > 0  24  22  1  1  2.315841E+01
  BBerr Tice(1) > 0 =  -6.680855917751967E-002
  BBerr Tice(2) > 0 =    12.3366102286845    
BBerr - Bug: IceT(2) > 0  25  22  1  1  1.158245E+01
BBerr - Bug: IceT(1) > Tmlt   4  23  1  1 -3.478426E-02 -5.400000E-02
  BBerr Tice(1) > 0 =  -9.795083681829635E-002
  BBerr Tice(2) > 0 =    33.7172473735695    
...

Note that the STDOUT errors only occur in two tiles (3 and 6, where 
the ice is).  The trial is seen in
/net/ds-01/scratch-4/enderton/GridTests/Coupled/c28_eq_ex_edge/results/Fail1/
where the STDERR and STDOUT files are in the Atm/STD directory.

Daniel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-devel/attachments/20050706/e989624d/attachment.htm>