On Sun, 14 Sep 2014, Loganaden Velvindron wrote: > Hi All, > > I would like to know if there would be any objectionn to having a > --status flag which would be useful to show logged in users. w and who > already do that. The output would go further: show list of forwarded > ports. Apart from --long_options being frowned upon, there is no single place that collects all this information. All sshd instances that are handling connections are completely* independent. There are arguments for and against making the master listener sshd take a more active role in the lifecycle of its children. The main for doing so is that it would allow us to better notice abnormal termination of sshd processes. E.g. noticing when waitpid() indicates that child crashed or the connection was closed for too many authentication attempts. Either of these could be useful signals for blackholing bad origins. The main argument against is that it builds more complexity into perhaps the most critical section of sshd code. If there is a bug in the master accept loop that causes sshd to crash, then users can no longer log in via ssh... -d * well, after authentication concludes (cf. MaxStartups) _______________________________________________ openssh-unix-dev mailing list openssh-unix-dev@xxxxxxxxxxx https://lists.mindrot.org/mailman/listinfo/openssh-unix-dev
