Changes between Version 1 and Version 2 of Archer


Ignore:
Timestamp:
15/01/16 17:55:03 (4 years ago)
Author:
annette
Comment:

—

Legend:

Unmodified
Added
Removed
Modified

  • Archer

    v1 v2  
    1010* /nerc: The Research Data Facility archive for long term storage. The RDF also has a dedicated processing system - the Data Analytic Cluster (DAC).  
    1111 
    12 For more information about the ARCHER service see the webpage: http://www.archer.ac.uk 
    13  
    14 '''To get an n02 account on ARCHER, contact grenville.lister AT ncas.ac.uk'''.  
    15  
     12For more information about the ARCHER service, including the hardware and software available see the webpage: http://www.archer.ac.uk 
    1613 
    1714== The UM on ARCHER == 
    1815 
    19 UM versions installed on ARCHER: 6.1, 6.6.3, 7.3, 7.8, 8.2, 8.4, 8.5, and 8.6. This means that you can run umui jobs on ARCHER for these versions of the model. However, we are still testing various model configurations; results of our tests can be found [wiki:Archer/HectorArcherComparison here]. 
     16UM versions installed on ARCHER are: 6.1, 6.6.3, 7.3, 7.8, 8.2, 8.4, 8.5, 8.6, 10.1, 10.2 and 10.3.  
    2017 
    21 === Setting up UMUI jobs for ARCHER === 
     18For all UM versions, whether using the UMUI or Rose, runs are submitted via PUMA.  
    2219 
    23 ''Things to do on PUMA'' 
     20=== Getting started ===  
    2421 
    25 Copy a HECToR UMUI job, name it appropriately, then set the machine name to login.archer.ac.uk - different versions of the model use for "machine name" various phrases; Host name in UM 8.5 for example or Other machine name in UM 6.6.3.  
     22To get an n02 account on ARCHER, contact grenville.lister AT ncas.ac.uk.  
    2623 
    27 ARCHER has 24 processors(cores) per node, so in order to use whole nodes (your job reserves whole nodes even if the number of cores required is not a multiple of 24), you should ensure that you use a multiple of 24 processors (a HECToR node had 32 processors). ARCHER has significantly more memory per processor than did HECToR, so if you were running with a non default number of cores per node, it's worth experimenting with this. 
    28   
     24Some steps are required before running a UM job for the first time on ARCHER:  
    2925 
    30 ''Version specifics'' 
     26* First register for a PUMA account by emailing the [ContactUs CMS team].  
     27* Set up your environment on PUMA and ARCHER by following [https://puma.nerc.ac.uk/trac/UM_TUTORIAL/wiki/UmTutorial/SettingUp these instructions].  
     28* For UM 10.x versions, see the instructions on [RoseCylc running the UM with Rose].  
    3129 
    32 For UM version 7.3 set the '''User machine overrides''' file to be 
     30=== Further instructions ===  
    3331 
    34 {{{ 
    35  ~umui/overrides/archer_cce_7.3_machine 
    36 }}} 
     32More information on specific topics can be found below:  
    3733 
    38 and the '''User file overrides''' file to be  
    39  
    40 {{{ 
    41 ~umui/overrides/archer_cce_7.3_file 
    42 }}} 
    43  
    44 ''Things to do on ARCHER'' 
    45  
    46 Set up ssh keys - you can simply copy the authorized_keys file from HECToR to ARCHER. 
    47  
    48 Ensure that your .profile is appropriate on ARCHER, again, this can probably just be copied directly from HECToR. Ensure the following lines are in your ARCER profile 
    49 {{{ 
    50 . /etc/profile 
    51 . /etc/bash.bashrc 
    52 }}} 
    53  
    54 Note that the environment variable TARGET_MC is set to cce on HECToR - this tells the UM build system to use the Cray compiler. On ARCHER we have again adopted the Cray compiler and TARGET_MC=cce remains the correct choice on ARCHER, however, we are experimenting with the Intel compiler. Initial tests indicate that there is little to choose between Intel and Cray. 
     34* [wiki:Archer/NercArchiving Setting up automatic archiving of UM output] 
     35* [wiki:Archer/DDT Debugging the UM on ARCHER with DDT] 
     36* [wiki:Archer/SCM Running the SCM on the ARCHER post-processor] 
    5537 
    5638 
    57 === UM Performance and domain-decomposition considerations === 
    5839 
    59 A quick scan of our test results indicates that jobs running on ARCHER with the same domain decomposition as on HECToR will run in half the wall clock time. However, the AU cost of an ARCHER core-hr is 15AU whereas a HECToR core-hr was 4AU, thus maintaining the same processor count will result in a cost increase by a factor of 15/8 , ie nearly twice as AU-costly on ARCHER as on HECToR. This represents good value since the NERC AU award is nearly 15/4 times bigger than on HECToR. However, it is well worthwhile experimenting with processor decomposition to find an efficient configuration. 
    60  
    61 === UM Archiving on ARCHER === 
    62  
    63 Instructions for setting up automatic archiving of UM output can be found [wiki:Archer/NercArchiving here]. 
    64  
    65 === Porting a job from MONSooN to Archer === 
    66  
    67 Instructions can be found [wiki:Faq/ConvertMonsoonJobToArcher here]. 
    68  
    69 === Interactive UM debugging with DDT === 
    70  
    71 Instructions can be found [wiki:Archer/DDT here]. 
    72  
    73 === Running the SCM on the ARCHER post processor === 
    74  
    75 Instructions can be found [wiki:Archer/SCM here]. 
    76  
    77 === Upgrading to cce8.3.7 === 
    78  
    79 Notes and instructions [wiki:Archer/cce8.3.7 here]. 
    80  
    81