weird ssh bug

weird ssh bug

Post by anth.. » Thu, 06 Aug 1998 04:00:00



Yesterday I tried connecting to one of my linux (x86) machines via ssh
and received the following error

Command terminated on signal 1.

and it kicked me right out.

After about a half hour of ripping out hair, I discovered that it was
caused by the fact I had the IdleTimeout set for 3d.
When I changed it to 2d or 4d everything was fine again.  So I left it
at 4d.  Then today I tried connecting again and got the same error.  So
I changed it back to 3d and everything worked fine.  So the next thing I
did was set the system date to 8/31 with a timeout of 30d and the same
thing happened.  Then to test the pattern

on 9/30/98 an IdleTimeout of 60d
on 12/31/98 an IdleTimeout of 152d
on 03/31/99 an IdleTimeout of  242d
on 04/30/99 an IdleTimeout of 272d
.....i gave up testing after that...

I tried this experiment on 8 different x86 machines running linux and it
happened on all of them.

so anyone that uses the idletimeout in the sshd_config and gets that
error..you know what to change.
Even though it was kind of annoying not being able to get into my
account remotely at least it didnt exploit a security hole.  This
problem happens on both redhat and slackware installations.  I tried
running sshd on a sparc ultra-1 with solaris os to recreate the problem,
but it didnt happen, so it might be an x86 thing.  If anyone has linux
installed on any other hardware platform (ppc, alpha, sun)  and has sshd
installed or is willing to install it, it would be intersting to find
out whether or not this problem will happen on those hardware platforms.

I tested this using ssh-1.2.22 which is the current version

http://www.cs.hut.fi/ssh/#current-version

 
 
 

weird ssh bug

Post by Frank Sweetse » Thu, 06 Aug 1998 04:00:00



> I tested this using ssh-1.2.22 which is the current version

> http://www.cs.hut.fi/ssh/#current-version

actually, 1.2.22 is pretty out of date, and has some serious security bugs
- instead, grab 1.2.26 from ftp.replay.com

--
Frank Sweetser rasmusin at wpi.edu fsweetser at blee.net | PGP key available
paramount.res.wpi.net RedHat 5.1 kernel 2.1.111p1  i586  | at public servers
Love makes the world go 'round, with a little help from
intrinsic angular momentum.

 
 
 

weird ssh bug

Post by anth.. » Thu, 06 Aug 1998 04:00:00


ok, I downloaded it and tested it out and that problem doesn't happen.

I guess that web site doesn't keep their pages current anymore.  I thought it
was weird that a new version hadn't shown up for a very long time on that page.

Thanks for the info.
:)



> > I tested this using ssh-1.2.22 which is the current version

> > http://www.cs.hut.fi/ssh/#current-version

> actually, 1.2.22 is pretty out of date, and has some serious security bugs
> - instead, grab 1.2.26 from ftp.replay.com

> --
> Frank Sweetser rasmusin at wpi.edu fsweetser at blee.net | PGP key available
> paramount.res.wpi.net RedHat 5.1 kernel 2.1.111p1  i586  | at public servers
> Love makes the world go 'round, with a little help from
> intrinsic angular momentum.

 
 
 

1. Weird Ping, weird FTP, weird Telnet... HELP!!!

Can anyone diagnose this for me?

I'm in the process of building up a firewall but at some stage I musta
messed up and am now getting really weird results.

This is the results of a ping from 192.168.0.1 to 192.168.0.1 (the box
which will become the fireawall).

# PING 192.168.0.1
PING 192.168.0.1 (192.168.0.1) from 192.168.0.1 : 56(84) bytes of
data.
64 bytes from 192.168.0.1: icmp_seq=0 ttl=225 time=0.7ms
64 bytes from 192.168.0.1: icmp_seq=1 ttl=225 time=80001.0ms
64 bytes from 192.168.0.1: icmp_seq=2 ttl=225 time=160000.9ms
64 bytes from 192.168.0.1: icmp_seq=3 ttl=225 time=240000.8ms
64 bytes from 192.168.0.1: icmp_seq=4 ttl=225 time=320000.8ms
64 bytes from 192.168.0.1: icmp_seq=5 ttl=225 time=400000.7ms
64 bytes from 192.168.0.1: icmp_seq=6 ttl=225 time=480000.7ms

^c

--- 192.168.0.1 ping statistics ---
509 packets transmitted, 8 packets received, 98% packet loss
round-trip min/avg/max = 0.7/280000.7/560000.6 ms

This also happens when I do it from other machines behind the firewall
but I was very surprised when it did it when pinging itself.

Theres a pattern of 80000ms increments between successful pings...
Can't figure out whats wrong though.

If I try to telnet or FTP, I have to wait for an eternity b4 the login
prompt appears after the initial "connected" msg. But once I login
there is no problem using the console or transferring files. Same
thing for other machines to 192.168.0.1. So the network is "working"
but its waiting for something.

Has anyone had this happen to them b4 and manage to solve it?

Any help appreciated

Regards,

2. KDeveloper sample?

3. Weird, weird, weird issue ....

4. netbeans port and jpda

5. can ssh-agent work between ssh-2.3 and ssh-3.5

6. Repartitioning for NT installation

7. Weird console / ssh lags in 2.4.x?

8. W shows "-"

9. weird ssh issue (security)

10. Weird "route add" issue with PPP over ssh linux 2.2.15

11. Ipmasqadm and ssh, weird problem

12. samba much slower than ssh and behaving weird??

13. Weird xfree86 bug