Hi all,
we're using an actor system in our play application with the default
strategies. One of our actor misbehaved and throw a StackOverflowError
which shutdown the actor system. The JVM was still running but the actor
system was down. I was wondering if it would make sense to catch
Hi Alexandre,
unfortunately it does not work like that. Recovering from SOE and OOME is
not possible in general, and as such it is better to fail fast and fix the
bug instead of limping along in a possibly corrupted state.
On Wed, Mar 25, 2015 at 3:13 PM, Alexandre Russel alexan...@russel.fr
In particular the concurrent data structures (including thread pools) might not
react well to VirtualMachineErrors, it is possible for them to just completely
lock up if you catch the error. Better recreate properly than stay silently
broken.
Regards,
Roland
25 mar 2015 kl. 16:10 skrev
