I'm trying to zero in on the exact option, or combination of options, that will force each SSH session to be terminated when inactive for a specified time. I'm using Putty to SSH to a RHEL AS3 server; keepalives are set to "off" (0) for Putty, and keepalives are commented-out in the ssh_config and sshd_config files.
I've Google-bounced to different places, and I've found some tidbits on ServerFault, but I'm hoping to find a "concrete" solution. I've tried modifying the /etc/ssh/sshd_config file by adding this:
ClientAliveInterval 15 ClientAliveCountMax 1
(It's set at 15 seconds for testing)
Then restarted SSH:
/sbin/service sshd restart
But no luck; the session never timed out (checked back after 23 minutes). I also tried modifying the "keepalive" kernel parameters:
% echo "60" > /proc/sys/net/ipv4/tcp_keepalive_time (default was 7200)
% echo "1" > /proc/sys/net/ipv4/tcp_keepalive_probes (default was 9)
% echo "5" > /proc/sys/net/ipv4/tcp_keepalive_intvl (default was 75)
But the session remained active (checked back after 8 minutes). Any help on what piece of the puzzle I'm missing would be greatly appreciated. Thanks!
--- DT