getpid?

[Tom Spence_1]

Hi All,

I read the file 'syslog.log' and wonder what is it?

telnetd[10786]: getpid : peer died: No such file or directory

Thank you for your help.

Tom

[Sridhar Bhaskarla]

Looks like the client side of it died abruptly. It happens mostly on the PCs where the user might have improperly closed the windows or the session.

Is it appearing too often?. It could be because of the connectivity issues also.

-Sri

[Paula J Frazer-Campbell]

Hi

This normally occurs on a dropped telnet session where the user has not done a normal exit.

Most of it is from the little "x" on a Windoze session window where a user just presses "x" instead of "exit" from the prompt.

A correct "Exit" allows Unix to tidy up correctly whereas an "x" from the window just drops the connection and Unix reports that event and the fact that it is cleaning up.

HTH

aula

[A. Clay Stephenson]

This is not a big deal. It simply means that the process on the other end of a telnet connection can no longer be reached. Telnetd, the telnet daemon, runs on the server enter and answers requests by telnet sessions. It could be as simple as someone telnetting in using their PC and powering down the PC (or possibly one of those very rare, almost unheard of, PC operationg system crashes.

[Tom Spence_1]

Very interesting about this... Ok, is there any way for me to fix it?

Tom

[S.K. Chan]

Well not so much of how to fix it but rather how to find out which client is causing the TCP connection termination. How many these entries you see in syslog.log and how frequent ? What you can do is restart inetd with "-l" (logging) if you have not already use "-l" in inetd. That way any client that attempt to make a telnet connection will get logged and you would know the ip addr of that client that causes thet getpid error.

[Tom Spence_1]

I checked with inetd.conf and it said;

telnet stream tcp nowait root /xxx/tcpd /etc/telnetd -b /etc/issue

It isn't often happened... Just once a while...

Tom

[A. Clay Stephenson]

The way you fix this is to stop PC clients from crashing; God might be able to do that but I doubt that you will have complete success.

[S.K. Chan]

Not "telnetd" but "inetd" ..
# ps -ef|grep inetd
==> Does have "-l" ?
# cd /etc/rc.config.d
# grep INETD netdaemons
==> Is INETD_ARGS set to "-l" ? That way it'll make sure inetd is started with "-l" option. You said "once a while" .. if I were you I wouldn't waste my time digging in further. It's normal. I do see a couple in my syslog.log for the past 1 month.

[Anil C. Sedha]

Tom,

I agree with Clay and others who specified earlier that it is a dropped telnet session.

Many users just exit out of their sessions without typing exit and kill the window instead. The telnetd daemon sees this as a broken pipe and hence errors out in the syslog file as a peer connection failure or dropout.

The error pertains to that.

Regards,
Anil (It is always a good thing to assign points when you find the answers valuable to you)

[doug hosking]

This is really not something to lose any sleep over. The message is ugly, and perhaps disturbing if you don't know why it happens, but I agree with the others. It's almost certainly because the connection was dropped abruptly by the client, such as by closing
a window without exiting the telnet session first. You can safely ignore it.
I'll ask the owners of HP-UX's telnetd if they will consider making the message a bit less cryptic to reduce needless confusion/concern in the future.