Jim Jagielski wrote:
On Aug 18, 2006, at 9:58 AM, Jean-frederic Clere wrote:
Jim Jagielski wrote:
I and other have run into issues where the socket
between Apache and Tomcat (due to a in-between firewall)
isn't closed as it should be...
Yep, I have heared about things like that, on the Tomcat side the
socket seems opened but it closed on the httpd side.
The problem is that httpd will detect it and close the "closed"
socket but nothing reaches the TC... Until a timeout occurs on the
TC side (OS socket timeout).
New requests cause new connections to be opened that results in
threads increase (x2) in the TC with all the bad you could think of.
Exactly.
I'm digging further into
this as far as why the timeout isn't being honored, but
it got me thinking that a "no reuse" option might be
nice. Basically, it prevents reuse from ever being set
to TRUE...
Well the problem is how to detect that a "bad close" has happended.
Yep. Code-wise it's not trivial (not real tough, not
just trivial) but from an Admin standpoint, they now when
their setup causes this. So having a "don't reuse" flag
would allow them to bypass this.
Right I haven't find the right way to handle this until now.
I have developped a JMX interface over the httpd-proxy-scoreboard
slotmem that allows TC to mark its worker not useable when the number of
threads increases too much. Somehow don't use until I am ready.
That would allow the Admin of the TC to handle httpd workers.
The other way I see it to force a PING/PONG when starting a new
connection, there TC could say don't use until I am ready (with a time
value).
I am only -0 about the "don't reuse" ;-)
Cheers
Jean-Frederic
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
