[Nagiosplug-devel] [ nagiosplug-Patches-3358348 ] check_smtp SIGPIPE (rc=141) error fix

SourceForge.net noreply at sourceforge.net
Fri Jul 15 13:17:23 CEST 2011


Patches item #3358348, was opened at 2011-07-08 08:22
Message generated for change (Comment added) made by duncan_ferguson
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=397599&aid=3358348&group_id=29880

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Bugfix
Group: release-1.4.15
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Duncan Ferguson (duncan_ferguson)
Assigned to: Nobody/Anonymous (nobody)
Summary: check_smtp SIGPIPE (rc=141) error fix

Initial Comment:
When running check_smtp against some Exchange server versions the server can close the connection before the QUIT command is sent, resulting in a 141 return code (due to SIGPIPE signal).  This patch will ignore the SIGPIPE and catch the error when it occurs.

This appears to be a non-fatal error as no other problems have been reported about the Exchange server by SMTP users and there are no problems identified in the Exchange logs.

$ ./check_smtp -V
check_smtp v1.4.15 (nagios-plugins 1.4.15)
$ ./check_smtp -H 192.168.0.1  -v
HELOCMD: HELO localhost
220
Connection was closed by server before we could send QUIT command
SMTP OK - 0.001 sec. response time, ping
|time=0.000984s;;;0.000000
  rc = 0


----------------------------------------------------------------------

>Comment By: Duncan Ferguson (duncan_ferguson)
Date: 2011-07-15 11:17

Message:
Added a patch to make the action configurable.

Default action is to exit unknown with a message  This is overridable by
using '-q' so the error is ignored (but still seen when using -v)

----------------------------------------------------------------------

Comment By: Ulrik Haugen (ulrik_haugen)
Date: 2011-07-13 09:15

Message:
check_smtp returning 141 is obviously a problem, but masking the cause of
it and returning STATE_OK instead seems an even bigger problem to me.

The server closing the connection prematurely could be a resource problem.

If this happens for you so often that tuning max_check_attempts,
normal_check_interval and retry_check_interval doesn't help then a patch
for check_smtp may be needed, but in that case ignoring that the connection
closes unexpectedly must be and option.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=397599&aid=3358348&group_id=29880




More information about the Devel mailing list