Project

General

Profile

Actions

action #18920

closed

[tools][sles][functional] test fails in apache_ssl - Serial output is written but not detected by os-autoinst

Added by SLindoMansilla almost 7 years ago. Updated about 6 years ago.

Status:
Resolved
Priority:
Normal
Assignee:
-
Category:
Bugs in existing tests
Start date:
2017-05-03
Due date:
% Done:

0%

Estimated time:
Difficulty:

Description

Observation

openQA test in scenario sle-12-SP3-Server-DVD-ppc64le-minimal+base+proxy_SCC@ppc64le fails in
apache_ssl

Serial output is written but not detected by os-autoinst

osd#912085/file/serial0.txt

02:07:47.5370 47625 <<< testapi::type_string(string=' ; echo pnPuk-$?- > /dev/hvc0
', max_interval=250, wait_screen_changes=0)
02:07:50.6790 Debug: /var/lib/openqa/cache/openqa.suse.de/tests/sle/tests/console/apache_ssl.pm:20 called apachetest::setup_apache2
02:07:50.6791 47625 <<< testapi::wait_serial(regexp=qr/(?^:pnPuk-\d+-)/, timeout=90)
02:10:25.9505 47625 >>> testapi::wait_serial: (?^:pnPuk-\d+-): fail
02:10:25.9507 47625 # Test died: command 'ls /etc/apache2/ssl.crt/$(hostname)-server.crt /etc/apache2/ssl.key/$(hostname)-server.key' failed or timed out at /var/lib/openqa/cache/openqa.suse.de/tests/sle/lib/apachetest.pm line 56.

osd#912085/file/serial0.txt

UUIrp-0-
pnPuk-0-
_shae-0-

Reproducible

Fails since (at least) Build 0365 (current job)

Not reproducible cloning job into osd: osd#913244

Expected result

Last good: 0363 (or more recent)

Further details

Always latest result in this scenario: latest

Workaround

Restart the job

Actions #1

Updated by RBrownSUSE almost 7 years ago

  • Status changed from New to Feedback

Not reproducable, most likely a timing issue in the test or the SUT being overworked, not sure what the purpose of this ticket is or the work being proposed.

Actions #2

Updated by SLindoMansilla almost 7 years ago

  • Description updated (diff)

The purpose is to investigate the issue and add information. At least making the easy and quick way of retrieving data for reproducibility.

(It could be that you already know what is the problem and that the solution would be incrementing the timeout parameter, just hypothetical, and then you are seeing this ticket as useless)

But from my inexperience with such problems, I am not able to know so far if it is a product bug or test issue. An even if it is a test issue, this ticket is aimed to find a possible solution. In the worst case, someone with more experience can meta-review my work, like you did, and point me to the right direction.

I am doing my best ;)

Actions #3

Updated by okurz over 6 years ago

  • Target version set to Milestone 13
Actions #4

Updated by riafarov about 6 years ago

  • Status changed from Feedback to Resolved

Doesn't happen for multiple months on SLE 12, nor for SLE15 in last 3 months. I believe that actually command has timed out, but obviously we do not kill the process, so it was written to the serial.

Actions

Also available in: Atom PDF