root [rcanvas] Comment out SaveAs() in last tutorial

This is currently the only tutorial calling RCanvas::SaveAs() and it fails very often (more than 90% of the time) on fedora38 and fedora39 since the last update of chromium.

Feb 16 '24 13:02 hahnjo

Starting build on ROOT-performance-centos8-multicore/soversion, ROOT-ubuntu2204/nortcxxmod, ROOT-ubuntu2004/python3, mac12arm/cxx20, windows10/default How to customize builds

Feb 16 '24 13:02 phsft-bot

Build failed on ROOT-ubuntu2204/nortcxxmod. Running on root-ubuntu-2204-2.cern.ch:/home/sftnight/build/workspace/root-pullrequests-build See console output.

Failing tests:

Feb 16 '24 13:02 phsft-bot

Build failed on ROOT-ubuntu2004/python3. Running on root-ubuntu-2004-1.cern.ch:/home/sftnight/build/workspace/root-pullrequests-build See console output.

Failing tests:

Feb 16 '24 16:02 phsft-bot

Test Results

12 files 12 suites 2d 4h 28m 22s :stopwatch: 2 561 tests 2 560 :white_check_mark: 0 :zzz: 1 :x: 28 810 runs 28 808 :white_check_mark: 0 :zzz: 2 :x:

For more details on these failures, see this check.

Results for commit 1fa1b22c.

Feb 16 '24 22:02 github-actions[bot]

Hi @linev, I understand that disabling the test is not the ideal action, but this one is failing in 90% of the builds since a week and this is not acceptable. It is now also seen failing on Jenkins (https://github.com/root-project/root/pull/14755#issuecomment-1951712063) which disproves the claim that it's the machine configuration - on Jenkins, there aren't the warnings about too little space on /tmp. If you are unhappy about disabling this test, please investigate with critical priority how to make the test pass. If that cannot be done within a day or two, I maintain that we should disable the test and go back to a green CI state.

Feb 19 '24 07:02 hahnjo

@hahnjo

How can I reproduce this build:

https://lcgapp-services.cern.ch/root-jenkins/job/root-pullrequests-build/196057/consoleFull

I only see that is executed on olbdw-01 host, but no any docker/vm configurations.

Feb 20 '24 07:02 linev

@hahnjo

How can I reproduce this build:

https://lcgapp-services.cern.ch/root-jenkins/job/root-pullrequests-build/196057/consoleFull

I only see that is executed on olbdw-01 host, but no any docker/vm configurations.

Because that one is the "old" Jenkins, ie no container. There the old procedure applies, ie logging in via SSH. For the container based CI, there's instructions at the end of every build how to reproduce it.

Feb 20 '24 07:02 hahnjo

I download docker for fedora39 on my machine, rebuild ROOT and do:

ctest -R rline
Test project /github/home/ROOT-CI/build
    Start 1109: tutorial-rcanvas-rline.cxx
1/3 Test #1109: tutorial-rcanvas-rline.cxx .........   Passed    2.40 sec
    Start 1110: tutorial-rcanvas-rline_style.cxx
2/3 Test #1110: tutorial-rcanvas-rline_style.cxx ...   Passed    1.10 sec
    Start 1111: tutorial-rcanvas-rline_width.cxx
3/3 Test #1111: tutorial-rcanvas-rline_width.cxx ...   Passed    1.11 sec

100% tests passed, 0 tests failed out of 3

Feb 20 '24 07:02 linev

Also simply running rline.cxx macro in the docker also works.

Feb 20 '24 07:02 linev

Yes fine, but that doesn't change the fact that it is regularly failing in the CI environment

Feb 20 '24 08:02 hahnjo

One should try to return back shared memory config parameters for container.

Feb 20 '24 08:02 linev

We are running in circles... As Axel explained on Mattermost, they were never active. Also:

It is now also seen failing on Jenkins (#14755 (comment)) which disproves the claim that it's the machine configuration - on Jenkins, there aren't the warnings about too little space on /tmp.

Feb 20 '24 08:02 hahnjo

Some resources - /tmp, /home, memory - is not sufficient for the container. Can it be 16 parallel jobs which are used to run ctest? Can one mark some macros not to run in parallel?

Feb 20 '24 08:02 linev

As Axel explained on Mattermost, they were never active

I was not aware about this.

Feb 20 '24 08:02 linev

Can it be 16 parallel jobs which are used to run ctest? Can one mark some macros not to run in parallel?

https://github.com/root-project/root/pull/14771; I'm not really convinced that's a solution, but worth a try...

Feb 20 '24 08:02 hahnjo