----------------------------------------------------------------
BEFORE YOU POST, search the faq at <http://java.apache.org/faq/>
WHEN YOU POST, include all relevant version numbers, log files,
and configuration files. Don't make us guess your problem!!!
----------------------------------------------------------------
> From: Marc Slemko[SMTP:[EMAIL PROTECTED]]
> Subject: Re: Many JServ threads being made and dying.
>
> I'm not sure I understand what the problem is or why anything you describe
> is a problem.
>
> TIME_WAIT is a normal TCP state that things go through during the
> shutdown of a TCP connection. There is nothing "stuck" about it and
> just because there is a connection in TIME_WAIT doesn't mean there
> is some process waiting for it.
Every time I execute "netstat" there are a couple (two or three) TIME_WAIT
state TCP connections that have dissapeared and have been replaced with
fresh new ones.
> jserv's use of threads requires that a new one is created and destroyed
> for each request; on Linux, these will show up in a normal ps because
> of how Linux threads work.
But five threads a minute are dying (and being replaced) WHEN THE SERVER IS
IDLE AND NO REQUESTS ARE BEING MADE. Besides, I thought there was a thread
pool being used so that threads were reused.
> jserv automatically connects to the JVM to verify that it is up
> and running and happy. This is not done if you start the JVM
> manually, since then it is your problem to ensure it is running
> properly.
>
> This seems to explain everything you see without there being any
> "problem".
The "problem" is that the behavior of Apache JServ as a server process
differs depending on whether I (a) let JServ start up the JVM as is the
default configuration or (b) I start the JVM manually.
If I start the JVM manually, I do NOT have over 300 JVM threads created and
destroyed every hour regardless of whether any requests have been processed
and I do NOT constantly have 10 or so TCP connections timing-out in the
TIME_WAIT state.
This is one of those particularly frustrating problems because it doesn't
really seem to cause noticable operational problems with the server
(although I can't guess if there might be some subtle side-effects) so it
has managed to go on this long ignored.
Murray
> --
> --------------------------------------------------------------
> Please read the FAQ! <http://java.apache.org/faq/>
> To subscribe: [EMAIL PROTECTED]
> To unsubscribe: [EMAIL PROTECTED]
> Archives and Other: <http://java.apache.org/main/mail.html>
> Problems?: [EMAIL PROTECTED]
>
--
--------------------------------------------------------------
Please read the FAQ! <http://java.apache.org/faq/>
To subscribe: [EMAIL PROTECTED]
To unsubscribe: [EMAIL PROTECTED]
Archives and Other: <http://java.apache.org/main/mail.html>
Problems?: [EMAIL PROTECTED]
