The bug seems to come randomly and actually some days doesn't appear at
all. Just some minutes ago the bug appeared again. I had a look at the
process information and saw the memory consumption rise up to 1 GB
before manually stopping the service. This is as expected for endless
recursion.

As that's a bad situation in case other mission-critical apps are
running on the same server, perhaps you should think about a generic
solution for endless recursion in this place. I know similar situations
from PHP and C projects and it is really simple to implement an
emergency brake by counting method entrance (decreasing on exit).
Stopping the recursion at a certain level (let's say 10 recursions)
would usually avoid server crashes and let the user try to press the
non-responsive key again.

Beyond that, maybe it is possible to store a stack trace and other
debug information then, so you'll finally be able to find out what
happens there and who's involved.


-- 
Shnapoo
------------------------------------------------------------------------
Shnapoo's Profile: http://forums.slimdevices.com/member.php?userid=39489
View this thread: http://forums.slimdevices.com/showthread.php?t=30908

_______________________________________________
beta mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/beta

Reply via email to