Re: who flags on OSR5


----- Original Message -----
From: "Jeff Hyman" <scolist@xxxxxxxxxx>
Newsgroups: comp.unix.sco.misc
To: <distro@xxxxxxx>
Sent: Tuesday, March 21, 2006 3:54 PM
Subject: who flags on OSR5


# uname -X
System = SCO_SV
Release = 3.2v5.0.6
KernelID = 2000-07-27
Machine = Xeon

Does anyone have any ideas why the 'who' command would fail to show a
logged in user with command:

# who -f

yet the command:

# who -x
shows the user.
User is logged in on a ttyp*
--


I have no idea what the -f is supposed to do, or why. I read the man page
but that still doesn't clarify anything.
(what is a superfluous, non remote, pseudo tty user? who -h adds
superfluous to what the man-page says)
On a box with about 220 users, all but 5 remote in terms of being not within
the netmask of any nic, only one non-remote in terms of tcp vs serial or
console, who -f shows exactly one line, and it's not even myself (ssh on
ttyp50) but a console session on tty01.
Checking a few other smaller boxes shows almost the same thing.
Sometimes my own ssh session shows up, sometimes nothing at all is retruned
even though 20 or so mixed lan & wan users are logged in.

The boxes in question have users all on ttyp*, using a mix of facetwin and
openssh.
Point being that it's no particular server login daemon doing something odd
(or failing to do something) in utmp that who fails to parse.

I just tried rlogin from a remote box (via vpn) in to the 220 user box and
who -f shows my own ttyp55 of the rlogin session and the previously noted
tty01.

gwho (/usr/gnu/bin/who, if you have gnutools installed) does not appear to
have any analogous option.

Whatever it's supposed to do, I see no use for what it actually does.
I never used who -f but I do use who and who -ux and who -umx a lot.
who -umx is also part of a script I run in /etc/profile on all my boxes.

It's also always been true that who and w are only in sync with reality as
long as the various login daemons correctly update /etc/utmp & friends.
There are various ways utmp can find itself incorrect. I don't think this
issue is what you're after though. It's somewhat rare and the symptoms are
different anyways.
They (w, who) are only, and should only be used as, a convenience that is
usually right, not an authoritative/definitive source that can be 100%
relied on.
ps is the real authority.

--
Brian K. White -- brian@xxxxxxxxx -- http://www.aljex.com/bkw/
+++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++.
filePro BBx Linux SCO FreeBSD #callahans Satriani Filk!

.

Relevant Pages

  • Re: "tsadmin" cannot show user list
    ... "tsadmin" need 30 min to show user list. ... Because, command "quser/qwinsta" are new seach form internet, I will ... "tsadmin" have remote control user's session function. ...
    (microsoft.public.windows.terminal_services)
  • Re: Remote Printing from a DOS app
    ... the other two are remote offices. ... command apply to all users regardless of their rights or if they are power ... Windows programs print to the remote computer fine. ... > I even wrote a script to do that automatically as the session ID changes ...
    (microsoft.public.windows.terminal_services)
  • Re: How do I find out who is connected via Remote Desktop to an XPP machine?
    ... Thanks for pointing out this useful command. ... When somebody is remote desktopping to that machine, ... "administrator" against that session (which is the local user name I have ... connecting from - which is what I am after). ...
    (microsoft.public.windowsxp.work_remotely)
  • Re: Remotely query local policies
    ... RSOP.MSC is a GUI on boxes XP and newer that you can remote to another box ... but I am not sure about a command line option. ... I have admin access and I do not have access to push/pull ...
    (microsoft.public.windows.server.general)
  • [UNIX] GNU Mailutils Multiple Vulnerabilities (Buffer Overflows, Format String, DoS)
    ... execute arbitrary code from remote using the GNU Mailutils programs. ... Remote exploitation of a format string vulnerability in the imap4d server ... The vulnerability specifically exists in the handling of the command tag ...
    (Securiteam)