ZenPacks : Nagios Check Ping : Comments

RE: Nagios Check Ping

community@zenoss.org — Wed, 09 Dec 2009 20:28:25 GMT

I've had this plugin sucessfully installed for sometime and after applying the template to a device the graph shows the error: MISSING RRD FILE: <servername> ipcheck_RTA

Thoughts?

RE: Nagios Check Ping

community@zenoss.org — Wed, 23 Dec 2009 06:22:18 GMT

Found the Template is using a base path of /usr/local/zenoss/libexec. Does anyone know where this base path is defined? (NOTE: The Redhat package I installed placed itself into /opt/zenoss but some of the compiled pyc/o files have references to /usr/local/zenoss - seems like a packaging issue)

There's probably a better fix, but I modified the Data Source to explicitly use /opt/zenoss/libexec/check_ping and it now works.

Steps:
Settings->ZenPacks->ZenPacks.Nagios.CheckPing
Then select /Devices/rrdTemplates/NagiosCheckPing

Select the ipcheck DataSource and prefix check_ping with /opt/zenoss/libexec/

Should be good to go after that - just double check that the change propagates to your devices.

RE: Nagios Check Ping

community@zenoss.org — Wed, 23 Dec 2009 15:29:18 GMT

Ok, investigated some more and the problem seems to be due to the Redhat distribution.

The base class /Device has the zCommandPath set to /usr/local/zenoss, even though zenoss was installed to /opt/zenoss. I changed the zProperties:zCommandPath for /Device to /opt/zenoss. [I'm also going to look to see if I can fix just /, if that can be corrected, that would be the better fix as everything else inherits from /]

After restoring the original ipcheck DataSource to use just "check_ping" and checking the DeviceClass and device zProperties were using zCommandPath of /opt/zenoss - it worked.

I believe this is the correct fix.

RE: Nagios Check Ping

community@zenoss.org — Mon, 01 Feb 2010 13:26:24 GMT

Don't see like a RedHat problem cause I'm on Ubuntu and have the same problems. In the Ubuntu case, the value of the zCommandPath zProperties that solve my problem was "/usr/local/zenoss/common/libexec/", of course... in the /Device class. Seem like a bug, cause the value that have before the change was a directory that don't exist: "/usr/local/zenoss//libexec/".

RE: Nagios Check Ping

community@zenoss.org — Fri, 05 Feb 2010 05:34:32 GMT

On my 2.5.1 virtual appliance version of zenoss this zenpack returns the following error during installation

Output:

zenpack --install /tmp/ZenPacks.Nagios.CheckPing-1.0-py2.4.egg.zip

ERROR: zenpack command failed. Reason: OSError: [Errno 20] Not a directory: '/home/zenoss/Products/ZenPacks.Nagios.CheckPing-1.0-py2.4.egg/skins'

Done installing ZenPack.

Anyone else have this problem?

RE: Nagios Check Ping

community@zenoss.org — Fri, 05 Feb 2010 08:14:41 GMT

Ignore the above. I forgot to unzip the zenpack before trying to install it from the GUI

RE: Nagios Check Ping

community@zenoss.org — Wed, 30 Jun 2010 08:54:16 GMT

I'm running Zenoss Core 2.5.2 on Debian Lenny, I installed this zenpack following the procedure and I'm experiencing the same problem as described above (I changed the zCommandPath to /usr/local/zenoss/common/libexec/check_ping which is the right path on my system)

the command runs fine when invoked from the shell and gives :

PING OK|LOSS=0 RTA=18.85 |rta=18.851999ms;180.000000;300.000000;0.000000 pl=0%;100;100;0

The "Ping - RoundTripAverage" graph is empty and shows the message "MISSING RRD FILE: <device> ipcheck_RTA"

The "Ping - Loss" is missing, no graph and no errors displayed, just a "missing image" icon is displayed

Any idea why it doesn't work as expected ?

Thanks

RE: Nagios Check Ping

community@zenoss.org — Fri, 09 Jul 2010 14:27:50 GMT

Ok Fixed it - the problem is that the check_ping command rely on the DNS name of the device, which can differ from the device name used in the GUI. I created an entry in the /etc/hosts file so it resolves accordingly, and the graphs started to show

RE: Nagios Check Ping

community@zenoss.org — Mon, 30 Aug 2010 19:53:22 GMT

I have 2.5.2 on Debian Lenny, so far I have no problem with check_ping via DNS name.

However, using it on zenoss 3.0.1 seems to have the above problem. I have many zombie processes with check_ping IP in UNIX ps that device is using IP address.

Is anyway check_ping can use IP instead of "-H $devname"?

RE: Nagios Check Ping

community@zenoss.org — Wed, 01 Sep 2010 14:16:06 GMT

I found this "-H ${device/manageIp}" instead of "-H $devname" that was using DNS name.

RE: Nagios Check Ping

community@zenoss.org — Tue, 05 Apr 2011 13:08:15 GMT

I guess this is a ZenPack that isn't evolving anymore but still, it's working on Zenoss 3.1.0

Everything is okay except by default PingRTA graph is showing abnormal values (i had spikes up to 300seconds on the graph!)

After reading some forum posts i came to a conclusion that there is a mistake in zenpack's default GraphPoint for Ping RTA graph.

Go to the NagiosCheckPing performance template, open "Ping - RoundTripAverage" graph definition (Manage Graph Points) and edit the "RTA" data point FORMAT filed:

change

%5.2lf%s

to 

%5.2lfms

This will correct the value displayed on the graph so instead of e.g. 300s I now get 300ms, and that's more like it!

marko

RE: Nagios Check Ping

community@zenoss.org — Tue, 13 Sep 2011 17:54:12 GMT

Some helpfull tips:

With Zenoss 3.1.0 I was given an error message while trying to install:

Link to http://pypi.python.org/simple/ZenPacks.Nagios.CheckPing/ ***BLOCKED*** by --allow-hosts

Couldn't find index page for 'ZenPacks.Nagios.CheckPing' (maybe misspelled?)

Link to http://pypi.python.org/simple/ ***BLOCKED*** by --allow-hosts

No local packages or download links found for ZenPacks.Nagios.CheckPing==1.0
Couldn't find index page for 'ZenPacks.Nagios.CheckPing' (maybe misspelled?)

No local packages or download links found for ZenPacks.Nagios.CheckPing==1.0
ERROR:zen.ZenPackCmd:zenpack command failed

I solved this by renaming the zenpack from "ZenPacks.Nagios.CheckPing-1.0-py2.4.egg" to "ZenPacks.Nagios.CheckPing-1.0" and it installed fine.

Also followed mvlaovic's tip and changed the RTA data point format.

As well as reading what trgrimm did, mine didn't work straight off the bat. My check_ping was located in a different location. ('/usr/local/zenoss/common/libexec/'). So I changed the Command Template to '/usr/local/zenoss/common/libexec/check_ping' from 'check_ping'.

How I found my check_ping was with: (With /usr/local/zenoss where I had installed zenoss)

find /usr/local/zenoss -name check_ping

RE: Nagios Check Ping

community@zenoss.org — Tue, 13 Sep 2011 18:09:47 GMT

This is good info. David, could you export your setup as a ZenPack so that others could easily use this functionality in zenoss 3.x? All 3.x zenpacks have py2.6.egg at end of file name, because they work on python version 2.6, which is default in zenoss 3.x

RE: Nagios Check Ping

community@zenoss.org — Tue, 13 Sep 2011 21:39:37 GMT

Hey,

I'll most certainly do so. Also another tip I've just learned, when I was attempting to get the Zenpack to run on another VM (Running CentOS 5.4) compared to my VM with CentOS 6.0. I found that the sed command that the check_ping pipes to, produces the following error:

Exception while performing command: 
Traceback (most recent call last):
  File "/opt/zenoss/Products/ZenUI3/browser/command.py", line 108, in stream
    data = unjson(request.form['data'])
  File "/opt/zenoss/Products/ZenUtils/jsonutils.py", line 175, in unjson
    return _json.loads(value, **kw)
  File "/opt/zenoss/lib/python2.6/json/__init__.py", line 318, in loads
    return cls(encoding=encoding, **kw).decode(s)
  File "/opt/zenoss/Products/ZenUtils/jsonutils.py", line 37, in decode
    result = super(StringifyingDecoder, self).decode(s)
  File "/opt/zenoss/lib/python2.6/json/decoder.py", line 319, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/opt/zenoss/lib/python2.6/json/decoder.py", line 336, in raw_decode
    obj, end = self._scanner.iterscan(s, **kw).next()
  File "/opt/zenoss/lib/python2.6/json/scanner.py", line 55, in iterscan
    rval, next_pos = action(m, context)
  File "/opt/zenoss/lib/python2.6/json/decoder.py", line 183, in JSONObject
    value, end = iterscan(s, idx=end, context=context).next()
  File "/opt/zenoss/lib/python2.6/json/scanner.py", line 55, in iterscan
    rval, next_pos = action(m, context)
  File "/opt/zenoss/lib/python2.6/json/decoder.py", line 155, in JSONString
    return scanstring(match.string, match.end(), encoding, strict)
ValueError: Unterminated string starting at: line 1 column 164 (char 164)

So, if you receive this error, I think I tracked down the issue to the sed versions on 5.4 being 4.1.5 while my newer CentOS uses 4.2.1, how I just managed to resolve this issue is the by changing the sed portion to:

replace " - Packet loss = " "|LOSS=" | replace "%, RTA = " " RTA=" | replace "ms|" "|"

However, it is not pretty but it does however work.

RE: Nagios Check Ping

community@zenoss.org — Thu, 25 Oct 2012 20:31:01 GMT

In Zenoss v.4.1.1 I had to modify the sed portion of the command to the following:

sed -e 's# - Packet loss#|LOSS#;s#%,#\;#;s#ms##;s# = #=#g;s#|rta=.*##'

PING OK|LOSS=0; RTA=0.73

RE: Nagios Check Ping

community@zenoss.org — Tue, 12 Nov 2013 15:32:02 GMT

Hi All,

Anyone has the idea on how-to make check_ping work with Zenoss on "CRITICAL - Packet Filtered" errors?