[might want to view this in a monospaced font] I am running ssh 1.2.26 on Solaris 2.5.1 and 2.6 and I keep getting these in console & messages file: Mar 24 14:04:26 www sshd[6646]: fatal: Local: Bad packet length 1077952576. Mar 24 14:05:27 www sshd[6836]: fatal: Local: Bad packet length 1077952576. ... large scp's and rsyncs (over ssh) seem to freeze between two machines on different sides of a network. Machines on the same segments have no problems. The network between the two machines who are talking looks like this: 100BT 6Mbps over OC3 100BT ------[Positron]================[Positron]------------- | | +----------+ +----------+ | gate | | jammer | | firewall | | firewall | +----------+ +----------+ | 10BT | 100BT +-------+ +-------+ | waldo | <--source destination--> | www | | 2.6 | | 2.5.1 | +-------+ +-------+ We used to have the two firewalls on the same segment, and this worked fine then. This leads to my own personal theory that maybe something on the Positrons is changing the packets from an MTU for 100BT to a 6-Mb MTU (and back again?), perhaps screwing up something along the way? (That might explain why it only happens when copying larger files.) Can someone tell me whether I might be on the right track on this? I am going to run some tests from different parts of the network when I get a chance, and maybe change the MTU on the interfaces to the lowest common denominator. Thanks for any help anyone can give me. --James # James Welborn (JW1560) <[EMAIL PROTECTED]> # Mac lover and Unix system administrator # Webmaster, the new Vegas.com <http://www.vegas.com/> # Nevada's Largest Website
