inetd Going down on signal 13

[jeffrey groff]

What exactly does this mean and what are the possible causes of this?

[Brian Hackley]

Hi,

/usr/include/sys/signal.h defines the signals:
define _SIGPIPE 13 /* write on a pipe with no one to read it */

I went to one of my HPUX 11 test systems and did a kill -13 on inetd, but it ignored the signal. Maybe you can run inetd -b (undocumented debug mode) then inetd -l to start connection logging, and watch syslog until you have another failure??

I don't what what OS nor inetd patch you have, and you didn't indicate if this happens "all the time" or just happened once. So I don't know if I helped or not...

-> Brian Hackley

[linuxfan]

Hi Jeffrey,

Are you seeing these messages in the /var/adm/syslog/syslog.log file or does this error message show up when you do something?

-Regards
I am RU

[jeffrey groff]

Sorry for the lack of details. I'm on 10.20 and it only happened once. The error is being reported in syslog. I looked up signal 13 in signals.h just like you did and found the "pipe" explanation. I'm not going to worry about unless it becomes a chronic problem. Was just curious.....

[Sridhar Bhaskarla]

Something might have caused inetd to create a pipe where obviously it wrote but there was no process to read. Since it is a broken pipe, kernel sent a signal 13 might be to delete the pipe from it's tables.