Use different exit status for provisionning errors and test errors

Bug #1133227 reported by Jean-Baptiste Lallement
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
UTAH
Fix Released
Medium
Javier Collado

Bug Description

A common pattern in test jobs is:

run_utah_test [arguments]...
RC=$?

if RC != 0; then
    case $RC in
        1) do_something
        2) do_something_else
        *) do_something_with_unhandled_codes
else
   all_good

UTAH seems to return the same code for different type of issues, especially provisioning and test errors. This is problematic in the case of provisioning error when the rest of the job script expect that the test environment is up but is actually not.

Expected Behaviour:
- Use specific exit code to identify provisioning errors and test errors.

If it's already the case, then it should be documented.

Tags: ue-desktop

Related branches

Revision history for this message
Javier Collado (javier.collado) wrote :

Looking at the code I've seen that the server returns some codes depending on
the failure or, if everything works fine, it returns what it got from the
client.

The potential problem here is the overlapping of return codes from the client
and the server. It looks like "1" can be returned both from the client and the
server on different failures. Hence, the server caller is unable to figure out
if the problem happened at the client or at the server without looking at the
logs.

tags: added: ue-desktop
Changed in utah:
assignee: nobody → Javier Collado (javier.collado)
status: New → Triaged
importance: Undecided → Medium
Changed in utah:
status: Triaged → Fix Committed
status: Fix Committed → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.