Changes between Initial Version and Version 1 of Docs/CAP9.0InstallationArcher

09/02/16 13:10:27 (4 years ago)


  • Docs/CAP9.0InstallationArcher

    v1 v1  
     1= CAP9.0 Installation on ARCHER = 
     3On PUMA, the code was extracted from the Met Office Science Repository, using the command, 
     5fcm co fcm:ancil.x-tr -r 307 
     7The revision number corresponds to CAP9.0. 
     9The following three files were created in the fcm-make directory, 
     11 * ancil_ARCHER_cce_debug.cfg 
     13 * ancil_ARCHER_cce_opt.cfg 
     15 * inc/ARCHER-cce.cfg  
     17These were based on the Met Office XC40 equivalents. The changes related to directories on the local computer. The configuration files seem to impose the use of absolute directories. 
     19With these changes, the command, 
     21fcm make -f $HOME/CAP9.0/fcm-make/ancil_ARCHER_cce_opt.cfg --new 
     22> CAP_build.log 2>&1 
     24was used to mirror the code over to ARCHER. 
     26On ARCHER, 
     28cd $DATADIR/CAP9.0 
     29fcm make --new 
     32This built the code successfully.  A new branch was created in the Met Office Science Repository to contain the ARCHER variant. 
     34== Issues Running the code on ARCHER ==  
     36=== Script modification === 
     38The scripts AncilScr_name were modified, where name is one of: top, mask, vegfracsoil, ozone, smow, sstice, slt, aerosol, orog. 
     40These were modified by eliminating the decisions based on the HARDWARE variable. The decisions that would have been made on the Met Office XCE computer were selected. Also, the first three were modified for the Cray assign problem. 
     42=== PBS commands === 
     44The PBS commands for XC40 are not compatible with XC30. The ncpus=1 construct is not accepted in the XC30 PBS select command. This led to failure to launch and no helpful error messages were produced. The issues are the specification of the queue is required on the XCE but not on ARCHER. 
     46=== Broken Links in the Ancillary Tree === 
     48The test programs use XML scripts which extract data from the ancillary tree. For example, in the directory {{{ancil/data/parameters/IGBP_to_MOSES_accurate}}} there are links such as, 
     50latest -> /projects/um1/ancil/data/parameters/IGBP_to_MOSES_accurate/v2 
     52where v2 is in the same directory as latest. These links pointing to absolute paths are present on the Met Office computer. I have modified the few on ARCHER necessary to complete the tests but there are 759 in all. There seems to be no reason for this. 
     54=== Comparison of Computers === 
     56The CAP scripts contain the line, 
     58HARDWARE=$(uname -a) 
     60which is then used to determine subsequent actions. The following code snippet from the scripts was run on Met Office computers and ARCHER, producing the results in the table below. 
     64HARDWARE=$(uname -a) 
     66if [[ $HARDWARE = +(*xc*) ]] 
     68  echo Hello 
     70 echo Goodbye 
     73if [[ $HARDWARE = +(*Linux*) ]] 
     75  echo Hello 
     77 echo Goodbye 
     82||Met O PC ||   Met O XCE       || MONSooN (Met O XCM)  || ARCHER || 
     83|| Goodbye || Hello  || Hello || Goodbye || 
     84|| Hello || Hello || Hello ||   Hello || 
     86showing that the Met Office Crays are set up differently from ARCHER.