Line | |
---|
1 | This directory contains the python cluster acceptance test system code. |
---|
2 | |
---|
3 | The programs here are designed to be run as a complete suite as well |
---|
4 | as individually. The suite writes to a log file as well as standard out. |
---|
5 | |
---|
6 | To run all tests in sequence, do: |
---|
7 | python test_all.py [<logfile>] |
---|
8 | If no logfile is supplied, ./test.log is assumed. |
---|
9 | |
---|
10 | -------------------------------------------------------------------------------- |
---|
11 | |
---|
12 | Ensure that the following environment variables are set to correct values: |
---|
13 | |
---|
14 | PYTHON |
---|
15 | contains the python to execute, eg, 'python2.4', etc. |
---|
16 | |
---|
17 | PYTHONPATH |
---|
18 | contains a single directory path to the ANUGA code. Something like |
---|
19 | /nas/gemd/georisk_models/inundation/sandpits/rwilson/rwilson/branches/numpy |
---|
20 | (only needed by the test_test_all.py module) |
---|
21 | |
---|
22 | EQRMPATH |
---|
23 | if set, contains a single directory path to the EQRM code. If not set, |
---|
24 | the EQRM code is not tested. |
---|
25 | |
---|
26 | -------------------------------------------------------------------------------- |
---|
27 | |
---|
28 | To run these tests on a new cluster: |
---|
29 | 1. Modify the 'Cluster_Info' structure in test_utils.py to add the new cluster |
---|
30 | information. |
---|
31 | 2. Create a new 'machines_<name>' file in this directory that describes the |
---|
32 | cluster compute-node setup. Note that the 'machines_<name>' file doesn't |
---|
33 | appear to be used by *anything*. |
---|
Note: See
TracBrowser
for help on using the repository browser.