remsh anf ftp not functioning

[Mike_316]

Hey Gang,

I have two identical workstations running HPUX 10.20. One is able to accept telnet, rlogin, ftp and remsh commands in as well as going out. The other one is giving the message "connection refused" when attempting to telnet or ftp to the box, and simply faisl when someone else tried to remsh to the box, however the box can telnet and ftp out.

I have compared the two boxes and everything looks identical. The .rhosts is the same for root on both boxes, there is no hosts.equiv for either, the inetd.sec looks the same between the two, the services file and the inetd.conf files are the same.

Where else should I look on the box failing to accept ftp and telnet sessions?

Thanks!
Mike

[David Lodge · Accepted solution]

It sounds to me like inetd isn't running...

[Mark Greene_1]

first eliminate the obvious: is inetd running? are there messages in syslog.log (or any of the log files in /var/adm/syslog) that refer to ftp or inetd? are both port 20 and 21 listed in /etc/services? Have you checked the entries in /etc/ftpd/ftpusers? Is ftp setup to log in /etc/inetd.conf (-l option)?

HTH
mark

[someone_4]

Hi
try
ps -ef |grep inetd

see if inetd is running.

vi /etc/inetd.conf
check to make sure the services you want are not commented out. If they are .. uncomment them then run
inetd -c
inetd -c rereads the inetd.conf file.

Let us know what you get.
also just wondering you said you cant ftp or remsh to it but can you ftp or remsh from it?

Richard

[someone_4]

I was able to duplicate it

with telnet and ftp in
/etc/services
on host box commented out.

# telnet lupus
Trying...
telnet: Unable to connect to remote host: Connection refused
# ftp lupus
ftp: connect: Connection refused
ftp>

after uncommeted in
/etc/services
they both worked

[someone_4]

opps hit reply to fast

now with just
telet and ftp commented out in
/etc/inetd.conf

# telnet lupus
Trying...
telnet: Unable to connect to remote host: Connection refused
# ftp lupus
ftp: connect: Connection refused
ftp>

same thing
after uncommeted
then running
inetd -c

it worked

[Corthouts Carlo]

Hi,

You could also check the file
/etc/inetd.conf if there are
lines for telnet and rlogin.
/etc/services should also contain these lines.

The file /var/adm/inetd.sec
checks if your system is denied or allowed access.

You should also check how many sessions are already running on that system.

It is most likely to many sessions are already running and all pty's are used.
Increasing the pty's and npty's could resolve your problem.

First check how many pty's you have available with :

ls /dev/pty |wc -w
ls /dev/ptym | wc -w

FYI : 60 is default
In this case the number of pty's have never been increased

First you need to increase npty in the kernel to 512.
(or any number you desire)

second you need to create additional pty's in /dev

insf -d ptys -n 512 (=npty)
insf -d ptym -n 512 (=npty)

[Mike_316]

OK. It was too simple. The reason I could find nothing wrong with the files, was they were all correct.

Although the inetd was running, it had been started with the old files and thus, was reacting to the security and process settings from the old files.

Stopped and restarted inetd, and everything is fine now.

Thanks again for all the help!
Mike