On 21.12.2009 13:55, mt...@apache.org wrote:
Author: mturk
Date: Mon Dec 21 12:55:02 2009
New Revision: 892803
URL: http://svn.apache.org/viewvc?rev=892803&view=rev
Log:
Fix #48276. Just like with setting port to 0, if the resolve fails, mark the
worker as disabled
Modified:
tomcat/jk/trunk/native/common/jk_ajp_common.c
tomcat/jk/trunk/xdocs/miscellaneous/changelog.xml
Modified: tomcat/jk/trunk/native/common/jk_ajp_common.c
URL:
http://svn.apache.org/viewvc/tomcat/jk/trunk/native/common/jk_ajp_common.c?rev=892803&r1=892802&r2=892803&view=diff
==============================================================================
--- tomcat/jk/trunk/native/common/jk_ajp_common.c (original)
+++ tomcat/jk/trunk/native/common/jk_ajp_common.c Mon Dec 21 12:55:02 2009
@@ -2575,6 +2575,13 @@
jk_log(l, JK_LOG_ERROR,
"worker %s can't resolve tomcat address %s",
p->name, p->host);
+ p->s->port = p->port = 0;
+ if (JK_IS_DEBUG_LEVEL(l))
+ jk_log(l, JK_LOG_DEBUG,
+ "worker %s contact is disabled",
+ p->name);
+ JK_TRACE_EXIT(l);
+ return JK_TRUE;
I'd say there's no easy way to decide whether that's intended (worker
with names prepared for deployment but not yet existing) or fatal (DNS
down, typo in worker name). Until know we decided to let this be fatal.
It's a big change to now simply ignore it.
If we want to support pre allocation of workers with names not yet
reolvable I'm fine with a global setting to allow that during startup,
but I'm not so happy to let everyone else run into issues because of
config typos.
WDYT?
Regards,
Rainer
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org
