| 16 | | $ ./configure <ROMS root path> |
| | 23 | $ perl --version |
| | 24 | |
| | 25 | This is perl, v5.8.8 built for cygwin-thread-multi-64int |
| | 26 | Copyright 1987-2006, Larry Wall |
| | 27 | |
| | 28 | }}} |
| | 29 | |
| | 30 | 2. Compile and install the required XML libraries.[Required][[BR]] |
| | 31 | The XML interface for the test framework requires Libxml2 (XML C parser and toolkit) as well as Perl bindings for libxml2.[[BR]] |
| | 32 | The complete step-by-step installation instructions for libxml2 and Perl modules, please go to [wiki:Libxml2-install Libxml2-installation].[[BR]] |
| | 33 | [Make sure you install xml libraries before proceeding further]. |
| | 34 | |
| | 35 | 3. Checks for availability of various compilers/utilities, creates workspace etc. |
| | 36 | * '''Perl'''[Required]: Version 5.8.8. |
| | 37 | * '''PBS utilities'''[Recommended]: Utilities such as qsub, qdel etc. are required to perform job scheduling. |
| | 38 | * '''mpirun'''[Recommended]: ROMS-CSTM parallel runs. |
| | 39 | * '''NCO tools'''[Optional]: The netCDF Operators, or NCO tools, such as ncdiff, ncks etc. |
| | 40 | * '''Workspace'''[Required]: Workspace to store test results. |
| | 41 | |
| | 42 | Go to the directory where you have checked out the beta version of the regression package. In subsequent documentation, we will refer the[[BR]] |
| | 43 | regression package home directory as '$REGRESSION_HOME. Type |
| | 44 | {{{ |
| | 45 | $ ./install_it -romspath <ROMS root path> |
| 21 | | Configuration script checks for availability of various utilities (perl, ncotools, libxml, qsub etc) and updates path and relevant information in various files.[[BR]] |
| 22 | | Provide a new/existing workspace name, when prompted to create a workspace for test results.[[BR]] |
| 23 | | '''You must run 'Configure before proceeding further.''' |
| | 50 | It is possible that your system may not have some optional/recommended softwares (as mentioned above), and the[[BR]] |
| | 51 | script may report that as warnings. It is highly recommended to install the optional softwares in order[[BR]] |
| | 52 | to take advantage of all the functionalities of regression test package. |
| 55 | | Check XML document ''''regressionCSTM.xml'''' in '/data/metadata' on how to configure model-run. Examples are also provided in '/data/metadata/example' to try.[[BR]] |
| 56 | | Examples 'example_simple.xml' which runs application 'UPWELLING' can be tried as it is without any further editing. Other examples may need some editing for |
| 57 | | directory path, etc.[[BR]] |
| | 85 | See a [wiki:sampleXML sample XML configuration] file. For further documentation, Check XML document in '$REGRESSION_HOME/data/metadata/regressionCSTM.xml'[[BR]] on how to configure model-run. Examples are also provided in '$REGRESSION_HOME/data/metadata/example' to try. The XML schema can be found in '$REGRESSION_HOME/data/metadata/schema/regressionCSTMSchema.xsd'. |
| | 86 | For a detailed schematic view of the schema, see [http://www2.msstate.edu/~skb12/nopp/regression/regressionCSTMSchema.html regressionCSTMSchema.xsd] |
| 59 | | Few good XML editors/validators are 'oXygen' and 'Architag's Xray'. |
| 60 | | |
| 61 | | [Detailed configuration help coming soon] |
| 62 | | |
| | 88 | == Try a Simple Test Example == |
| | 89 | 1. Open file '$REGRESSION_HOME/data/metadata/example/example_simple.xml' in any XML editor (oXygen or Xray). It defines a compilation and model[[BR]] |
| | 90 | run for application 'UPWELLING'. |
| | 91 | 2. Locate the tag '<environment>'. Edit the path information for various environmental variables and executables with reference to each compiler. |
| | 92 | 3. Locate the tag '<type>' nested within a '<compiler>' tag. There will be two '<type>' tags, one is set to 'ifort' and other is set to 'pgi'. [[BR]]This will compile and run test case 'UPWELLING' for 'ifort' and 'pgi' respectively. |
| | 93 | 4. Locate the tag '<modelRun>'. This ensures whether to carry out the model run or not. It will run the model (i.e. oceanS) from command line. [[BR]] |
| | 94 | If you would like to submit it via queuing system (i.e. pbs), then remove or comment the tag '<noPbs/>'. This will create a pbs script with [[BR]]specified processors and nodes and will submit a job to the pbs queue. The tags '<ppn>','<ncpus>' and '<nodes>' are all set to 1, for serial |
| | 95 | application 'oceanS'. |
| | 96 | 5. Save the file in the same directory. All xml documents has schema definition/document associated with them and needs to be placed in[[BR]] '$REGRESSION_HOME/data/metadata/example' directory for successful validation. Changing the schema reference path will be discussed later. |
| | 97 | 6. Go to '$REGRESSION_HOME'. We will now validate the conformity of edited XML document with the schema. Type |
| | 98 | {{{ |
| | 99 | $ ./bin/validate $REGRESSION_HOME/data/metadata/example/example_simple.xml |
| | 100 | }}} |
| | 101 | If there are no syntax errors and xml document conforms to the schema, then it will result in no errors. In case of any errors, check the error[[BR]] |
| | 102 | messages, edit the document, and repeat steps 5-6 till no errors. |
| | 103 | 7. Now, we will run the regression executable. Type (assuming you are in $REGRESSION_HOME) |
| | 104 | {{{ |
| | 105 | $ ./bin/runRegression -xml $REGRESSION_HOME/data/metadata/example/example_simple.xml |
| | 106 | }}} |
| | 107 | Based on the configuration defined in the 'example_simple.xml', the model will compile and run. |
| | 108 | 8. Once the test is finished, the test results will be placed in respective timestamped directory within the workspace [[BR]] |
| | 109 | which was created during the installation. See 'Understanding Test Results' below for more information. |
