TIME_WAIT/MSL is 2 seconds--bug/feature?

Joe Cooper (jcooper@pdq.net)
Tue, 04 Jan 2000 02:11:13 -0600


I've been having fits trying to figure this out, hopefully someone here
can help...

According to a small utility called msl_test distributed with the
Polygraph web cache benchmarking suite the Linux 2.2.13 TCP stack has a
TIME_WAIT of 2 seconds. As far as I can tell, the TCP specs recommend
60 seconds. Additionally, this is the required setting for IRCache
bakeoffs (and the only way to accurately compare to IRCache results). I
haven't checked any other kernels for this problem, but I have seen no
mention of it in numerous searches of the kernel mailing list.

Now, my confusiong lies in the fact that a grep through the kernel
source gives me this line in 'include/net/tcp.h':

#define TCP_TIMEWAIT_LEN (60*HZ) /* how long to wait to successfully
* close the socket, about 60 seconds */

Am I missing something, or shouldn't this set the TIME_WAIT value to 60
seconds? As far as I can determine, it does not. I there another place
I should be looking?

Does anybody know what's going on who can point me in the right
direction? I'd be very grateful for some elucidation.

Thanks in advance. Please CC to me, as I'm afraid I might miss it in
the digest.

-- 
Joe Cooper <joe@swelltech.com>
http://www.swelltech.com

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/