Issue with xinetd. Need assistance

[acidbox]

I checked my /var/log/messages this morning and I'm getting thousands of messages like this:

Nov 23 17:01:52 talon xinetd[7928]: service poppassd, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service submission, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service poppassd, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service submission, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service poppassd, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service submission, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service poppassd, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service submission, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service poppassd, accept: Invalid argument (errno = 22)
Nov 23 17:01:52 talon xinetd[7928]: service submission, accept: Invalid argument (errno = 22)

I've tried restarting xinetd and qmail with no resolve. The process looks like this:

# ps -aux | grep xinetd

root      7928  0.1  0.0  2148  928 ?        S    Nov22   2:16 xinetd -stayalive -pidfile /var/run/xinetd.pid
root     30944  0.0  0.0  2148  932 ?        S    17:00   0:00 xinetd -stayalive -pidfile /var/run/xinetd.pid

/var/run/xinetd.pid contains:

30944

30944 refers back to the process listed above. Beyond this, I have no clue what this means.

Does anyone know what this means and how to fix it?

 

[rvdmeer]

Check the file /etc/xinitd.d/poppassd_psa

On a fedora core 3 server it should look like this:

[root@vite xinetd.d]# cat poppassd_psa

service poppassd
{
socket_type = stream
protocol = tcp
port = 106
wait = no
disable = no
user = root
instances = 1000
flags = KEEPALIVE
server = /usr/local/psa/admin/bin/poppassd
}
[root@vite xinetd.d]#

 

[acidbox]

I'm on RHEL3 and mine looks like this:

service poppassd
{
socket_type = stream
protocol = tcp
port = 106
wait = no
disable = no
user = root
instances = 1000
flags = KEEPALIVE
server = /usr/local/psa/admin/bin/poppassd
}

Appears to be the same as yours. I haven't made any changes to anything related to this, and it's been running fine for months, which is why I find this error odd.

 

[rvdmeer]

There are 2 pids running?

Hmmm on a normal server ther is just one pid running. they might fight with eachother.. i would recommend killing the pid. The only thing is.. be sure you 're able to reboot the server with a powerswitch if something goes wrong with connections.

 

[acidbox]

Originally posted by rvdmeer
There are 2 pids running?

Hmmm on a normal server ther is just one pid running. they might fight with eachother.. if you restart the xinet process, what happens?

/etc/init.d/xinetd restart

...

I know. weird.

after restarting, i still show:

root      7928  0.1  0.0  2148  928 ?        S    Nov22   2:16 xinetd -stayalive -pidfile /var/run/xinetd.pid
root     32752  0.0  0.0  2152  940 ?        S    17:26   0:00 xinetd -stayalive -pidfile /var/run/xinetd.pid

I checked the messages log and I see that just before the error started, I got this:

Nov 22 11:38:15 edge xinetd[7928]: pmap_set failed. service=sgi_fam program=391002 version=2
Nov 22 11:38:15 edge xinetd[7929]: pmap_set failed. service=sgi_fam program=391002 version=2

And right before that, it appears that BIND restarted and qmail restarted.

 

[rvdmeer]

i corrected by post:

Hmmm on a normal server ther is just one pid running. they might fight with eachother.. i would recommend killing the pid. The only thing is.. be sure you 're able to reboot the server with a powerswitch if something goes wrong with connections

 

[acidbox]

do you think rebooting the server might fix it?

 

[rvdmeer]

We'll.. i think killing pid: 7928 will fix it as well.
But i'm not sure of that. And because when killing a pid on something network based you must be able to restore from that with a powerswitch. If your server is not connected to a powerswitch i would recommend rebooting it.

 

[acidbox]

Thanks for the help.

I killed the PID and then restarted xinetd. That seems to have fixed it.

I now get:

Nov 23 17:32:32 edge xinetd: xinetd shutdown succeeded
Nov 23 17:32:32 edge xinetd: xinetd startup succeeded
Nov 23 17:32:32 edge xinetd[564]: pmap_set failed. service=sgi_fam program=391002 version=2
Nov 23 17:32:33 edge xinetd[564]: xinetd Version 2.3.12 started with libwrap loadavg options compiled in.
Nov 23 17:32:33 edge xinetd[564]: Started working: 7 available services

with no errors after it.

 

[rvdmeer]

I think those 2 processes where fighting with eachother

 

[acidbox]

Not sure how that got started. One thing I see that looks like an issue is:

Nov 23 17:32:32 edge xinetd[564]: pmap_set failed. service=sgi_fam program=391002 version=2

Do you know what that means?

 

[rvdmeer]

There are several results when searching on google. Just check those pages to see what the problem is. I guess they will help you along.

http://www.google.com/search?hl=en&q=pmap_set+failed.+service%3Dsgi_fam+

 

[acidbox]

Thanks.

It looks like this post was the answer.

Basically my portmap service wasn't started.

service portmap start
service xinetd restart

That did the trick!

Thanks.