SSH Host Termination Times

[gary2]

When using SCO 5.06 Enterprise, SSH connections are terminated by the host when the connection drops for 5 seconds or so. This is either in-house or internet connections.

This message appears in the emulator debug tool:

[recv_ex got error 10054]
[WINSOCK errorerror 10054: An existing connection was forcibly closed by the remote host.]

Is there a way to increase the time the host tolerates for these dropped connections? Many users have less than perfect internet connections. This is not reffering to idle sessions.

Thanks,
Gary

 

[bbScripter]

I see two possible problems here, the ssh connection and/or the TCP/IP stack on SCO.

The sshd server has two values that control keep alives, stored in the sshd_config file:

ClientAliveInterval
ClientAliveCountMax

Check the man page for more detail, but these default to off. So its probably NOT your sshd server resetting the connections.

At the other end, you may want to check the clients ssh configuration. PuTTY at least has an option to send keepalives.

Its been a while for SCO, but a quick search on google turns up this man page:

http://osr5doc.sco.com:457/cgi-bin/man/man?tcp+ADMP

Which basically states that TCP keeps the session alive for like 511 seconds. But there are a few overrides that you can see within that doc.

I hope that gives you some ideas on where to look. Good luck.