-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Konstantin,
On 10/31/2011 12:19 PM, Konstantin Kolinko wrote: > 2011/10/31 Christopher Schultz <ch...@christopherschultz.net>: >> >> PS Is there anything to be done about connector state after an >> OOME? Chuck contends that OOMEs aren't certain death for an >> application and that recovery is possible. > > The problem is that with OOME you do not know what exactly remained > in the broken state. > > Generally, the thread that encountered any VirtualMachineError > will rethrow the error up the whole stack die. As far as I remember > my TC7 testing several months ago, for request processor threads it > is not fatal, as new processors will be created. The processors might be re-created, but some of the other shared, re-usable resources (buffers?) might be re-used and in some kind of broken state. In the case I observed, here, can you think of a condition under which the buffer would be left in a broken state and not re-set on a subsequent request? I wonder if Tomcat could be slightly more resilient under such circumstances. Thanks, - -chris -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk6u5qcACgkQ9CaO5/Lv0PAqqACdFf5n3/AtG99xOfRwkBfv7671 vqYAn2b73wjpcBqKpgALGtPTQMZ/CIAi =rOpJ -----END PGP SIGNATURE----- --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
