[Pkg-nagios-devel] Bug#739254: Bug#739254: nagios-plugins-basic: check_ssh causing remote servers to log Connection reset by peer messages.

[Jim Barber]

On 17/02/2014 5:21 PM, Jan Wagner wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Hi Jim,
>
> thanks for taking the time to report the issue.
>
> Maybe this is caused due the fix of #734811. But could you please
> describe the system that is affected?
>
> Thanks, Jan.
>

Hi Jan.

I'm not 100% sure what you are asking of me, but I'll try to answer your question.

We have an Icinga monitoring server that runs numerous checks against various
hosts around our network.
This is the host that I upgraded the nagios-plugin* packages on, and then had
to back them out again to the older version.
One of those checks run by the Icinga server is using the check_ssh plugin, and it
is that check that is causing the remote monitored hosts to log the error message.

The Icinga server is primarily running Debian 'testing' packages, but has a
couple of 'unstable' packages that didn't filter down to testing yet.

The remote monitored hosts are all Debian systems of various ages.
The versions of the openssh-server package that I've seen installed on these
hosts are:
    6.0p1-4
    6.2p2-3
    6.2p2-4
    6.4p1-2
    6.4p1-3

All of these are outputting the error when the Icinga server is checking their
SSH daemon via check_ssh, so it doesn't seem to matter what the version of the
remote SSH daemon is.

Many of these monitored hosts are on the same 10Gbit LAN as the Icinga server,
while some are across a WAN with a 10Mbit connection.
The speed of the link doesn't seem to have an impact on the results seen above.

Is there anything specific that you need to know that I haven't covered above?

Regards,

Jim Barber