telnet

[cute]

Hi, whew quite busy today... :-) what a day!



Im trying to telnet to my machine, i have 50 users logged in my system and my system cannot accept anymore telnet sessions. error msg: 'telnetd: telnet device drivers missing: No such device'



check how many users are logged on -> 50

checked my npty -> 60

checked nstrpty -> 60

checked nstrtel -> 60



does this mean i still have a margin of 10 [60 less 50 equals 10]?



but why cant my users log on?



please help..

Thanks!!!

[cute]

i run who -r and found only 50 telnet sessions..

does HPUX reserve telnet sessions?



i think i am losing 10 telnet session based on my computation... is that correct or is there something technical about this?

how do i know if the system uses some telnet sessions?



help...

Thanks

[Animesh Chakraborty]

Hi,
Check the process running by the user.
#ps -ef|grep user|wc -l
You may kill some hung user process for a quick remedy.

[Bill McNAMARA_1]

Seems like the device files are not created:
'telnetd: telnet device drivers missing: No such device'

ll /dev/pt* | wc -l

perhaps an insf -e or
insf -eCtty will create them.

Later,
Bill

[cute]

Thanks animesh,

but does that mean i will have to check each and evryone of my users?

i can see all of them and they are all active connection...

i am wondering about the npty=60 kernel parameter, I have just 50 telnet sessions so this means i haven't reached my 60 limit? but i can no longer telnet to my server...



THank you so much!!!

[Steven Sim Kok Leong]

Hi,

You may not have sufficient nstrpty slave-side stream special files for the pts driver. To create 60 slave-side special files (nstrpty default):

# insf -e -s

To create 80 slave-side special files:

# insf -e -s 80

Hope this helps. Regards.

Steven Sim Kok Leong

[Animesh Chakraborty]

Hi,
You can do rlogin to that machine.

I also get same messsages one of my server.
I my case what happens is every time my operators open omni screen and they never come out.Next time again they open a new omni screen.By doing this there are lot many xomni process keep on running on the server by the user "oper" and after few days we starts getting similar message.
After I kill all the xomni process it is back to normal.

I need to check increasing of npty value(in this case also it is 60) helps or not.

[cute]

Thanks for all the help guys!!!!

I plan to adjust my kernel config tonite.



But i still have one question, why cant i get any telnet session even though npty & nstrpty are set to 60 [default]? and i only have 50 active telnet connections...



does hpux use/reserve telnet sessions?

My boss is wondering about the missing 10 telnet sessions... hope you can give me details on this...



Thanks so much!!!!

[cute]

ooops, forgot to tell you guys...

i already executed 'insf -e'



will there be an error if i execute 'insf -e -s' while my users are logged in?





Thanks again!

[Steven Sim Kok Leong]

Hi,

Your kernel parameters may be set but the number of pty slave devices may not correspond in number. You will need to verify that via the above-mentioned methods.

If you check out the kernel parameters, it is stated for nstrpty that:

nstrpty should be set to a value that is equal to or greater than the number of PTY devices on the system that will be using streams-based I/O pipes.

For nstrtel:

nstrtel specifies the number of kernel data structures that are created at system boot time that are required to support the device files used by incoming telnet sessions on a server. This number should match the number of device files that exist on the system.

Hope this helps. Regards.

Steven Sim Kok Leong

[Steven Sim Kok Leong]

Hi,

Logically speaking, there should not be an error if you execute insf -e -s while your users are logged in because only more device special files will be created (if there were missing files previously) and not less.

Hope this helps. Regards.

Steven Sim Kok Leong

[cute]

Thanks everyone!!! i didnt give a 10 to avoid the rabbit to popup... sorry...





Thanks everyone!!!

[Animesh Chakraborty · Accepted solution]

Hi,
You need to change maxuprc to 200.
Default limit per user is 50.
And also change maxusers to 512.
You can run insf -e without affecting anyone.

[cute]

my problem was resolved by adjusting kernel parameters....



Now.... can i ask for another one?

I am having problems again.. upon loggin in

my server takes a long time to give a login prompt then logner time again for the password... then longer time again for the prompt...

Any idea what i should look for?



Thanks again...

[James R. Ferguson]

Hi:

If you are running DNS, the delay could be a reverse-name lookup timing. Make sure that your 'nsswitch' file specifies 'files' then 'dns' and try declaring your workstations in /etc/hosts even if you use dummy (bogus) names for their IP addresses.

Regards!

...JRF...