Summary

eliminate code in all tests that manages last_used_testplan; ; Owner: Tim ; Reporter: Tim ; Status: open ; Priority: high

; Summary: eliminate code in all tests that manages last_used_testplan;
; Owner: Tim
; Reporter: Tim
; Status: open
; Priority: high

Description [edit section]

= Description =
Currently, there is code in every single test's config.xml file (Jenkins
test configuration file), like the following:

{{{#!YellowBox
if [ ! -d "$FUEGO_LOGS_PATH/$JOB_NAME" ]; then mkdir -p "$FUEGO_LOGS_PATH/$JOB_NAME"; fi
        echo $TESTPLAN >$FUEGO_LOGS_PATH/$JOB_NAME/last_used_testplan;
        TESTPLAN=testplans/$TESTPLAN.json
}}}

This is followed by the actual invocation of the test base script using
a 'source' line.

It would be nice to eliminate this duplication and put these mkdir and file
creation operations into the functional.sh or benchmark.sh files, so they
don't clutter up each individual test.

Do some research to see why these are expressed this way, and if needed
refactor the handling of testplans to avoid this problem.  This looks
like it was done as a quick hack to save the last_used_testplan.

Also, why is the indentation in these files so weird?

Notes [edit section]

= Notes =
Also, why can't the 'source' line be calculated, and the entire config.xml
file generated at test package install time.  It probably could be if
the base script name were required to follow a convention (either have the
same name as the test, or have the name base_script.sh)

backlink

Fuego Issues List

; backlink: [[Fuego Issues List]]