Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations Mike Lewis on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

telnet connection refused

Status
Not open for further replies.

jmhdmc

MIS
Aug 31, 2001
2
0
0
US
I am setting up a new RS6000 eSeries 620 server and after configuring the tty ports, was able to connect via a modem and a terminal emulation client to the server. However, after installing the IBM / Informix Universe database, I cannot connect remotely. All connections are 'refused'. I can still ping the address and the console login is fine but not even the Universe Admin client can connect. I checked the user setting to allow remote logins, it is ok also.

I don't know what could have changed to disable remote logins. Any suggestions? Thanks in advance.

JJB
 
Check for the telnet port in /etc/services that it isn't commented out and check telnetd.
 
Hi,

Can telnet and the Universe Admin client be listening on the same port ? Use "netstat -A | grep LISTEN" to check which ports you are listening to.
 
Hi!
Rightnow do you try to connect the informix client over PPP protocal? Am i right?

 
Hi

I am also running IBM UniVerse on RS/6000 S80's and don't have this problem.(still on Aix 4.3.3 but hope to change in New Year). I'm interested in maintianing contact with another user of UniVerse on RS/6000. mail me at brian.mulvey@ie.flextronics.com if you like. Please no spam!

One thing to consider is that UniVerse has its own shell "uvsh". This normally lives in /usr/ardent/uv/bin (or wherever you installed universe. Check it is there and executable. Check if you Universe users are trying to use this shell in /etc/passwd

# ls -la /usr/ardent/uv/bin/uvsh
-rwxr-x--x 1 root staff 4746305 Jun 29 2001 /usr/ardent/uv/bin/uvsh

PTY count is the most obvious issue as stated above, but also does /etc/nologin exist?

if you check etc services for port 23 it should something like:

# cat /etc/services | grep 23/tcp
telnet 23/tcp
ntp 123/tcp # Network Time Protocol
cdc 223/tcp # Certificate Dist. Center
xinuexpansion3 2023/tcp
#
Inetd.conf should contain a line like:

# cat /etc/inetd.conf | grep telnet
telnet stream tcp6 nowait root /usr/sbin/telnetd telnetd -a
#


Make sure the telnetd program still exists:

# ls -al /usr/sbin/telnetd
-r-sr-xr-- 1 root system 417220 Jan 04 2002 /usr/sbin/telnetd
#


As far as I am aware the default port for the UniVerse rpc (which Admin uses ) is 31438. You can tune this but my guess is that you didn't set it to 23 (telnet)
To check run the command

# egrep uvrpc /etc/services
uvrpc 31438/tcp # uvNet rpc port


I hope some of this is helpful, and sorry for re-hashing what was already said.
 
Thanks to all.

I have been stepping through all the suggestions and have made some progress and am learning a few things. Now after stopping then restarting the tcpip services in AIX and RPC at the Universe level, I can connect but eventually get dropped and am back to square one with all connections refused.

This is progress but any more ideas would be greatly appreciated.

JJB
 
Maybe anothe couple of simple checks:

1.df - make sure nothing is 100% on the system volumes

2.Run:

mkpasswd -f

This can possible fix any password problems introduced by the Informix install.

3.Please post the output of errpt (apx. 10 lines) taken after the last connection failure. "Long live king Moshiach !"
h
 
Hi,
I assume you are trying to telnet to the server. Can you telnet to the Unix server via ip back to to the console? Can you telnet localhost? If you can, then it most likely is the subnet mask that is not allowing you to login. Does one subnet have a local ip and another a REAL ip? Did you check to see if you can telnet from another subnet that uses a different route altogether? (could be a permission problem on your gateway/router.)
If you cannot connect via telnet from the console, then we know the problem would be coming from the tcpip daemon, on the Unix server.
 
Did you run lsps -a to check on paging space used? If you are really low on paging space weird problems can happen because processes get killed of indiscriminantely.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top