Hello,
taking off from Michael's problem, we are also 
facing another problem, we are using JDK1.2 on Linux 6.0 with JSDK2.0 and
GNUJSP0.9.10
and ours is a web application where the development is on JSP technology.
Now we have to perform database queries on the server side depending
on the requests from the client side. 
We found out the following :
1. When we do a httpd start, about 11 process entries corresponding to httpd
are shown on
        doing a ps ax | grep http
2. For every database query that is fired, we get, as a sample
15696  ?  S    0:03 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j
15720  ?  S    0:00 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j
15721  ?  S    0:00 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j
15722  ?  S    0:00 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j
15723  ?  S    0:00 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j
15724  ?  S    0:00 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j
15725  ?  S    0:00 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j
15738  ?  S    0:00 /home/jserv/jdk1.2/bin/i386/native_threads/java
org.apache.j

3. Upto this point everything is fine. In my code I have a finally phrase
where I have 
a close() method written specifically for destroying any databse connection
associated with any particular class. It does simply rs.close() and
con.close() where
rs is of the type ResultSet and con is a Connection class object.

4. But I realise that even on destroy(), the entries in point 2. still
persist although
they should not have been there.
So this creates problems when multiple users are testing the application,
then we get arbitrary
databse connection errors. This, I think is the problem of native threads
being spawned, but
not SIGTERMed when the finally method is called.

5. Is the diagnosis given in point 4 correct ????
If so then what can be the solution for native thread problem ????
If the diagnosis is wrong, then what is the problem ???

Please let me know
Thanks and regards
Aditya


----------------------------------------------------------------------
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Reply via email to