rlogin hangs after 6.8 upgrade

[Whoever]

For what gain in a setup where security is completely irrelevant? Ssh needs key managment per machine and user and eats so many CPU cycles single-threaded that older computers cannot saturate a gigabit link, because you cannot disable this useless encryption.

This is a thread about a regression in 6.8, not about the advantages of ssh vs. rsh.

For the gain of getting away from using a protocol that almost nobody else cares about. Getting away from a protocol that was broken with a system upgrade.

[chemal]

RHBZ #1335670 has priority "urgent", severity "urgent", and now a fix. I don't care about your opinion.

[ernie]

A work around, if you are using xinetd, is to put nice = 5 in the rlogin config file in /etc/xinetd.d
Mine reads:

Code: Select all

service login
{
	socket_type		= stream
	wait				= no
	user				= root
	log_on_success		+= USERID
	log_on_failure 		+= USERID
	server			= /usr/sbin/in.rlogind
	disable			= no
	nice				= 5
}

There seems to be some race condition introduced by the later version of /bin/login very occasionally rlogin will work, by changing the nice level it seems to substantially increase the chance of success.