I think that's what we're experiencing here with Tomcat4.0.5...we have been
using TC401 and the shutdown is instant...but I think they have changed the
way TC is shutting down now...TC will gracefully shutdown i.e. waiting for
all threads finish and then exit...

If you want...you can try TC401, see if you still experience the delay...

I posted a similar message before and haven't got any response from the
community yet...:-(

Jindong.

 -----Original Message-----
From:   [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]

Sent:   Monday, February 10, 2003 10:19 AM
To:     [EMAIL PROTECTED]
Subject:        Shutdown.sh does not work when long lasting operations, such
as SQL Queries, are still active!


I already tried it on the tomcat-user list but my question problem/question
still remains. The problem
is as described below (and already posted in the tomcat-user list), that
tomcat
refuses to stop when long
lasting operations are still running. My primary concern is whether it is ok
to
start some extra threads from
within the Http processor threads. The J2EE specification strongly
discourages
the use of extra threads. Is there
any impact (besides resource consumption) on the runtime behavoir of tomcat
when
starting new threads
from the HTTP Processor threads?



Post on Tomcat-User:

Currently we are running a web application on AIX 4.3.3.0 with ibm jre 1.3.1
and
tomcat 4.0.3 final.
When trying to shutdown the tomcat server by means of shutdown.sh, we
encounter
the following problem:
The shutdown script is executed properly but the tomcat server does not
terminate. In my opinion this is
caused by long lasting operations within the http processor threads. In
other
words, when we have sql
queries or queries against other backend systems, like Content Manager (CM)
or
Enterprise Information Portal (EIP),
the tomcat server does not shutdown the context and call the destroy method
until all http processor threads
have terminated.

Is there a possibility to speed up the shutdown procedure, so that tomcat
does
not wait for active threads to come back?
If not is it a feasible solution to wrap long lasting operations in extra
threads and implement a time-out by means of the
Thread.join(long millis)  function within the http processor threads? Does
starting extra threads within the http processor
threads impose a problem to the runtime behavior of the tomcat server?


Thanks in advance,

Andreas



---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]


############################################################################
#########
This e-mail message has been scanned for Viruses and Content and cleared 
by MailMarshal
############################################################################
#########

#####################################################################################
This e-mail message has been scanned for Viruses and Content and cleared 
by MailMarshal
#####################################################################################

Reply via email to