On Sat, Sep 24, 2016 at 2:34 PM, Zbigniew Jędrzejewski-Szmek
<zbyszek(a)in.waw.pl> wrote:
That's a reflection on the nature of the change itself though. People
don't run large, stable workloads that would hit the TasksMax limit on
rawhide. They run them on stable releases. It's a chicken-egg
situation from a testing perspective, but if there is no announcement
or heads up the change is coming then people cannot know they should
test it before it hits stable.
josh
Although I agree with the above then may be a builder should have a rawhide VM running
with these updated critical packages if only for testing purposes to expose this type of
failure.
The other concerns is that the package should have issued some sort of critical error
message to the appropriate log before the limit was reached. and then start shutting down
or limiting the process's or threads running.
Is easy to say but may be a lot harder to actually achieve.
Although the change was communicated there needs to be a process that effective
communication of these sorts of changes is sent to ensure that it will be picked up by
more than just the fedora team leader which already has a fairly big workload.
It is a corner case at the extreme end as most testers probably would not have the
hardware or knowledge to test effectively.
Supplying info on how to test effectively may be a solution but will require someone to do
it.
But I feel Kevin's and others involved frustration and the amount of time that
trouble-shooting this would have taken.
Tim