Project

General

Profile

Actions

action #23712

closed

assert_screen seems to have stopped prematurely within VNC stall - test fails in install_and_reboot

Added by okurz almost 7 years ago. Updated over 6 years ago.

Status:
Rejected
Priority:
Low
Assignee:
-
Category:
Regressions/Crashes
Target version:
-
Start date:
2017-08-28
Due date:
% Done:

0%

Estimated time:

Description

Observation

openQA test in scenario opensuse-5.10.90-Argon-Live-x86_64-krypton-live-installation@64bit-2G fails in
install_and_reboot
with what looks like a stalled system.
What is interesting however is the following segment in the logfile:

20:53:09.6365 3181 no change: 71.6s
20:53:10.5435 3181 considering VNC stalled, no update for 4.00 seconds
20:53:10.6377 3181 no change: 70.6s
20:53:10.7935 3181 pointer type 1 0 1024 768 -257
20:53:10.7936 3181 led state 0 1 1 -261
20:53:11.6388 3181 no change: 69.6s
20:53:12.6321 3181 20:54:24.3288 3179 >>> testapi::_check_backend_response: match=rebootnow,screenlock,yast2_package_retry,yast2_wrong_digest timed out after 2000
20:54:24.8794 3179 <<< testapi::check_screen(mustmatch=[
  'yast-still-running',
  'linuxrc-install-fail',
  'linuxrc-repo-not-found'
], timeout=5)
20:54:25.0600 3179 >>> testapi::_handle_found_needle: found yast-still-running-buttons-20151113, similarity 1.00 @ 796/721
20:54:25.0602 3179 <<< testapi::select_console(testapi_console='install-shell')
20:54:25.2755 3179 <<< testapi::assert_screen(mustmatch='text-login', timeout=10)
20:54:36.3714 3179 >>> testapi::_check_backend_response: match=text-login timed out after 10
20:54:36.4220 3179 post_fail_hook failed: no candidate needle with tag(s) 'text-login' matched
20:54:36.4239 3179 test install_and_reboot failed

20:54:36.4248 3175 awaiting death of testpid 3179
20:54:36.4317 3175 test process exited: 3179
20:54:36.4318 3175 awaiting death of commands process
20:54:36.4348 3175 commands process exited: 3177
20:54:36.4349 3175 isotovideo done
no change: 68.6s
20:53:13.6333 3181 no change: 67.6s
20:53:14.6344 3181 no change: 66.6s
20:53:15.6356 3181 no change: 65.6s

It looks like either the internal method _check_backend_respond returned prematurely or the output with "no change 76.6s" and so on is delayed so that part of it's output only shows up after all the other blocks reported that the checks failed, including the next check for the install-shell as well as the handling of background process shutdown.

Reproducible

Fails since (at least) Build 1.99 (current job)

Expected result

Last good: 1.97 (or more recent)

Further details

Always latest result in this scenario: latest

Actions #1

Updated by okurz almost 7 years ago

  • Project changed from openQA Tests to openQA Project
  • Category changed from Bugs in existing tests to Regressions/Crashes
  • Priority changed from Normal to Low

as long as it's not seen elsewhere I regard this is low.

Actions #2

Updated by coolo over 6 years ago

  • Status changed from New to Rejected

I can't even spot a bug - the log output is not linear, but the timestamps give it away. So I see no road to action here

Actions

Also available in: Atom PDF