Wanting to add that the MTU workaround also used the
clear-dont-fragment-bit command, which is an ASP feature that is not
available on ordinary tunnel services pic.
http://www.juniper.net/techpubs/software/junos/junos63/swconfig63-interf
aces/html/interfaces-adaptive-services-config6.html
With this option you can avoid issues with the end-to-end application
never seeing an ICMP MTU too large error, and therefore retransmitting a
packet that is too large, while the default action of not allowing
fragmentation prevents successful communications.
Regards
-----Original Message-----
From: Harry Reynolds
Sent: Wednesday, November 30, 2005 7:02 PM
To: 'Terry Rodery'; 'juniper-nsp at puck.nether.net'
Subject: RE: [j-nsp] Adjusting MSS size for GRE tunnels in JunOS
AFAIK, not at this time. The TCP MSS is a system-wide
jnpr at foo> start shell
% sysctl -A | grep mss
net.inet.tcp.mssdflt: 512
net.inet.tcp.v6mssdflt: 1024
net.inet.tcp.send_mssdflt: 0
It is not recommended that folks mess with this type of
system default. I have heard that some folks have found a
work-around by increasing GRE tunnel MTU (say to 1600).
HTHs
-----Original Message-----
From: juniper-nsp-bounces at puck.nether.net
[mailto:juniper-nsp-bounces at puck.nether.net] On Behalf Of
Terry Rodery
Sent: Wednesday, November 30, 2005 1:51 PM
To: juniper-nsp at puck.nether.net
Subject: [j-nsp] Adjusting MSS size for GRE tunnels in JunOS
Hi,
I'm curious to know if JunOS on the M-series is capable of
doing that same as Cisco's "ip tcp adjust-mss 1436" on a
tunnel interface. I've done a bit of looking around and so
far I've only seen it mentioned for JunOS for ERX. Anyone
have any idea or could possible point me in the right direction? :)
Regards,
Terry Rodery
_______________________________________________
juniper-nsp mailing list juniper-nsp at puck.nether.net
http://puck.nether.net/mailman/listinfo/juniper-nsp