action #62624
closed[functional][y] Test if workaround for bsc#1041747 can be disabled
0%
Description
Motivation¶
We would like to test our product in the setup which is closer to the customer setups. Workaround for bsc#1041747 is to use single thread and single core.
We want to check if setup is stable enough to use it.
Updated by riafarov over 4 years ago
https://openqa.suse.de/tests/overview?distri=sle&version=15-SP2&build=128.1&groupid=129&arch=ppc64le didn't work well as we got 10GB disk for smp worker, checking if that also caused https://bugzilla.suse.com/show_bug.cgi?id=1161762
Updated by riafarov over 4 years ago
https://openqa.suse.de/tests/3825597#step/user_settings/3 shows that problem is still there. Reverting changes
Updated by riafarov about 4 years ago
- Status changed from In Progress to Resolved
Resolving, as we cannot remove the workaround. However, now issue is not with mistyping, but with meta-keys (alt, shift, etc.), so VNC_TYPING_LIMIT might help.
Updated by okurz about 4 years ago
There is also the possibility to select a different typing speed selectively, e.g. for the single line of code where typing fails. VNC_TYPING_LIMIT applies to the global test execution so that would slow down the tests potentially significantly but bsc#1041747 is really limited to "user_settings".
Updated by riafarov about 4 years ago
okurz wrote:
There is also the possibility to select a different typing speed selectively, e.g. for the single line of code where typing fails. VNC_TYPING_LIMIT applies to the global test execution so that would slow down the tests potentially significantly but bsc#1041747 is really limited to "user_settings".
On ppc we have seen these issues not only in that module, but also in bootloader. Actually, it's also visible with single thread and single CPU. As current plan is to drop power kvm, we are not considering to invest more time in it. However, scenarios which must be executed on multi threaded env, are already scheduled on those workers. So, based on this knowledge, we will keep things as they are. Jose is updated on this, so coverage for kernel tests will be adjusted accordingly if needed.