gmane.network.nagios.devel

Re: nagiostats Config Parameters

Jochen Bern — 2012-05-17T22:52:29

To answer my own (implied) question:

base/nagiostats.c uses a different subroutine, read_nagiostats_file(),
to read and parse a status file named by the -s option; when parsing
nagios.cfg (default or -c option) and grabbing the value of the
status_file directive there, read_status_file() is used instead.

As far as I can tell, read_status_file() matches the current status.dat
syntax, while read_nagiostats_file() expects a format that plain isn't
to be found anywhere else anymore. For example, it's the only place in
the code where I can find the string "nagios_version", which it parses for.

The figure nagiostats outputs for NUMSVCOK (the one I focused on in my
last post - *not* the only one resulting in wrong numbers) is actually
*counted* in status.dat by read_status_file(), while
read_nagiostats_file() expects it to be *stated* as "services_ok" in the
file read; again, this seems to be the only place in the code where
"services_ok" appears *as a string* (rather than a variable of the same
name).

Any cha

NDOUtils 1.5.1 Available

Eric Stanley — 2012-05-17T11:20:55

Version 1.5.1 of the NDOUtils add-on has now been released. You can 
download NDOUtils from:

https://sourceforge.net/projects/nagios/files/

Changelog follows...

       * Fixed off-by-one error packing data in ndomod.

Eric

nagiostats Config Parameters

Jochen Bern — 2012-05-11T19:34:39

Everyone,

I played with the "-c" and "-s" Parameters of a 3.2.3's nagiostats today
and noticed the odd behavior shown below. I didn't look at the source
yet, much less write a patch, but I thought you might want to check on
3.3.x and 3.4.x beforehand (and maybe extend t/705nagiostats.t a bit).

Kind regards,
J. Bern

Normal operation:

# N="/usr/local/nagios"
# $N/bin/nagiostats -m -d NUMSVCOK
1659


status.dat has been moved elsewhere (onto a ramdisk):

# strings - $N/bin/nagiostats | grep /nagios/
/usr/local/nagios/etc/nagios.cfg
/usr/local/nagios/var/status.dat
# grep /status.dat $N/etc/nagios.cfg
status_file=/usr/local/nagios/var/spool/status.dat
# strace $N/bin/nagiostats -m -d NUMSVCOK 2>&1 | \
open("/usr/local/nagios/etc/nagios.cfg", O_RDONLY) = 3
open("/usr/local/nagios/var/spool/status.dat", O_RDONLY) = 3


Specifying nagios.cfg explicitly works:

# $N/bin/nagiostats -c $N/etc/nagios.cfg -m -d NUMSVCOK
1659


Specifying status.dat explicitly - no go:

# $N/bin/nagiostats -s $N/var/spool/sta

Nagios Core 3.4.0 - performance data issues ifcontain apostrophes

2012-05-10T14:23:09

Hi list,

i'm having an odd problem with the development machine I installed  
Nagios Core 3.4.0 to.

It took me some time to make sure what the problem acutally is, but  
there must have some change in the performance data handling.

For my setup:
It's two machines running CentOS 5.8 x64, the production-machine is  
Nagios Core 3.3.1 and the development machine is Nagios Core 3.4.0,  
both have the same versions of other packages both have PNP-0.6.17 and  
both have nagios-plugins-1.4.15 - and both have the exact same config  
(except the dev-machine doesn't send email to everyone).

So far about the basics.

I'm having the problem with the check_snmp and check_nt scripts, if  
there is a label (on check_snmp) or a counter (on check_nt) with  
blanks and it's therefor in apostrophes.

So that means, that Nagios is still checking the state of the checks,  
but there is no performance data handled by PNP, as the input is  
invalid.

Let me explain that with an example:

This checks the environment temperature

Re: [Nagios-users] Any upcoming release?

Andreas Ericsson — 2012-05-09T21:43:27

Bleh. It's too late and to sit up and rebase 160-odd patches on top of
20 to fix a single broken one.

For those using the 3.4.0 release, I've cherry-picked all patches that
went in the 3.4.0-release apart from the broken execvp patch and added
them on top of the worker-process code.

If anyone's interested in trying them out, you can get them from
git://github.com/ageric/nagios.git or possibly get a snapshot somewhere
at https://github.com/ageric/nagios

Consider it 3.5.0-alpha or something. It still needs a lot of testing,
but it's been ticking along quite nicely in our lab the past 4 months,
running checks and generally just working with livestatus and Merlin
enabled.

Granted, that part of our lab doesn't use passive check results, but
most of the rest is tested and known to work pretty well, apart from
the small memory leak mentioned below (but who knows? Someone might
even fix that now that it's in the open).

Happy testing.

I'll get a proper fix in for the double-quote issue on friday, as I've
got pr

Re: Passive checks number of ettempts error

Rodney Ramos — 2012-05-09T12:50:12

Hi Stephen,

Thank you for your answer. However I don´t agree when you say that I´m
losing updates. We can see by the alert times that the collector sent a
SOFT3 status at
01:05:21 h and the central server consider it a HARD4 at 01:05:40 h. For me
it is a bug in the passive check process.

I´m using NSCA to replicate the status from the collectors to central
server and as you said the central server takes about 20 seconds to get the
status sent by the collectors, what is a normal behavior, I think.

Thank you very much,
Rodney

On Tue, May 8, 2012 at 5:01 PM, Stephen Gran <steve< at >lobefin.net> wrote:

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/__________________

Re: Passive checks number of ettempts error

Stephen Gran — 2012-05-08T20:01:41

On Tue, May 08, 2012 at 12:07:05PM -0300, Rodney Ramos said:

Hi,


You're losing updates.  Given that it seems to be taking 15 or 20
seconds to get the update from your collector to your central server,
that's not hugely surprising.  You don't say what the replication
mechanism is, but it either needs to get better at shovelling updates or
grow a bigger buffer, at a guess.

Cheers,

Passive checks number of ettempts error

Rodney Ramos — 2012-05-08T15:07:05

Hi everybody,

I use Nagios, release 3.2.3, in a distributed environment, with a central
server and several colector servers.

For a long time I´m seeing errors on the passive check mechanism on the
central server, as we can see below.

Sometimes, on the central server, the states and number of attempts don´t
follow the correct order, going from SOFT2 to HARD4, for example. However,
on the colector server everything is OK.

Log from Central Server:
Host Up[2012-05-05 01:06:48] HOST ALERT: node;UP;HARD;1;TCP OK - 0.005
second response time on port 135
Host Down[2012-05-05 01:05:40] HOST ALERT: node;DOWN;HARD;4;CRITICAL -
Socket timeout after 10 seconds
Host Down[2012-05-05 01:04:16] HOST ALERT: node;DOWN;SOFT;2;CRITICAL -
Socket timeout after 10 seconds
Host Down[2012-05-05 01:02:55] HOST ALERT: node;DOWN;SOFT;1;CRITICAL -
Socket timeout after 10 seconds

Log from Colector Server:
Host Up[05-05-2012 01:06:31] HOST ALERT: node;UP;SOFT;4;TCP OK - 0.005
second response time on port 135
Host Down[05-05-2012 01:

Re: Customized macros not working for contacts

Deborah Martin — 2012-05-04T15:44:34

Jochen, 

Thanks. Yes, I can create another template and add that as you suggest. I'd hoped to keep contacts defined within a contact definition and since it's documented that it should work .... 

But I'll try it out as you suggest and go from there. 

Regards,
Deborah 
Frustration=Finding something that you want to use documented only to find out it doesn't work! 

-----Original Message-----
From: Jochen Bern [mailto:Jochen.Bern< at >LINworks.de] 
Sent: 04 May 2012 16:02
To: Nagios Developers List
Cc: Deborah Martin
Subject: Re: [Nagios-devel] Customized macros not working for contacts

On 05/04/2012 04:28 PM, Deborah Martin wrote:

FWIW, you have a major systematic problem there. Event handlers are triggered *once* per service (and event), while said service can have
*any* number of contacts with their individual sets of Custom Macros. In your example, the service may have exactly one contact, but that's something Nagios would need to learn to check for to ever even get the idea that Contac

Re: Customized macros not working for contacts

Jochen Bern — 2012-05-04T15:01:38

FWIW, you have a major systematic problem there. Event handlers are
triggered *once* per service (and event), while said service can have
*any* number of contacts with their individual sets of Custom Macros. In
your example, the service may have exactly one contact, but that's
something Nagios would need to learn to check for to ever even get the
idea that Contact Custom Macros might be well-defined in your Event Handler.

Hence, I'ld even say that *Service* Custom Macros is where your
additional data *belongs*.

You are aware that you can reference *several* templates, as in "use
dbcon-service,special-contact-template", to combine perpendicular
default settings?

http://nagios.sourceforge.net/docs/3_0/objectinheritance.html
(Scroll to end of page)

Kind regards,
J. Bern

Re: Customized macros not working for contacts

Deborah Martin — 2012-05-04T14:28:51

Hi Paul, 

I don’t think that will be feasible. 

I have 10 template service definitions and these all contain static settings

Then for each customer, I define a service definition referencing that template and overriding only the parameters that are applicable to a specific customer such as the contact-group, 
host_name etc etc 

So for example, my template is defined below :- 

define service{
        name                            dbcon-service ; The 'name' of this service template, referenced in other service definitions
        active_checks_enabled           1       ; Active service checks are enabled
        passive_checks_enabled          1       ; Passive service checks are enabled/accepted
        parallelize_check               1       ; Active service checks should be parallelized (disabling this can lead to major
                                                ; performance problems)
        obsess_over_service             1       ; We should obsess over this service (if nec

Re: Customized macros not working for contacts

Paul Dubuc — 2012-05-04T13:01:06

Is is feasible to define the macro in a service template definition that your 
other service definitions use to get the value for the command-line parameter?

Deborah Martin wrote:


------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/

Re: Customized macros not working for contacts

Deborah Martin — 2012-05-04T12:30:11

Andreas, 

It doesn't work if I define a customized macro in the contact definition and try to use it at all! 

That means as a macro or as an environment variable. 

If I move the customized macro to my service definition then both work. 

The docs say this is supported so I believe this should be fixed. My configuration "is in the real world" and therefore would be very useful. As it's for use with an event handler I can't believe that it will have that much of a performance impact as the only time when the event handler needs to send out results is on a critical alert and that is based on certain criteria being met. 

Having the macro defined in the contact definition instead of the service definition is the difference between updating 1 contacts.cfg file which contain 30 different contacts versus updating 100's of service definitions where there is around 10 services for every contact. 

Regards,
Deborah 


-----Original Message-----
From: Andreas Ericsson [mailto:ae< at >op5.se] 
Sent: 04 May 2012 12:25
To:

Re: Customized macros not working for contacts

Andreas Ericsson — 2012-05-04T11:25:02

Use the macro as a command-line parameter from Nagios' config. From
what I read in your original mail, that seemed to work well, no?

Re: Customized macros not working for contacts

Deborah Martin — 2012-05-04T07:27:10

No - that was a typo on my part. 

In both test cases there were two underscores. 


Regards,
Deborah 

-----Original Message-----
From: eponymous alias [mailto:eponymousalias< at >yahoo.com] 
Sent: 03 May 2012 17:01
To: Nagios Developers List
Cc: Deborah Martin
Subject: Re: [Nagios-devel] Customized macros not working for contacts




Something is odd about the way you've written those names.
One has two underscores, the other has one.  Might that have something to do with the failure you see?


This e-mail and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed.  If you are not the intended recipient, please delete this e-mail immediately.  Any unauthorised distribution or copying is strictly prohibited.

Whilst Kognitio endeavours to prevent the transmission of viruses via e-mail, we cannot guarantee that any e-mail or attachment is free from computer viruses and you are strongly advised to undertake your own

Re: Customized macros not working for contacts

eponymous alias — 2012-05-03T16:01:25



Something is odd about the way you've written those names.
One has two underscores, the other has one.  Might that have
something to do with the failure you see?

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Nagios-devel mailing list
Nagios-devel< at >lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nagios-devel

Re: Customized macros not working for contacts

Deborah Martin — 2012-05-03T14:58:53

Thanks Andreas, 

Yes for both settings 

I do see your point about a fix. It's a shame because setting this in the contact definition would have made some event handlers easier to write. 

At the moment my event handler has to be customized per customer to get the right email contacts but if I could define a customized macro in the contact definition, I only need 
use that in my event handler script and then I don't need to worry about a different event handler per customer making it as generic as possible. 
Things like warning and critical thresholds are parsed based on customer requirements rather than hardcoded so I was hoping I could do that with contacts. 

Is there a different way of achieving this ? 

Regards,
Deborah 


-----Original Message-----
From: Andreas Ericsson [mailto:ae< at >op5.se] 
Sent: 03 May 2012 15:46
To: Nagios Developers List
Cc: Deborah Martin
Subject: Re: [Nagios-devel] Customized macros not working for contacts

On 05/03/2012 12:42 PM, Deborah Martin wrote:

Are environment macros ar

Re: Customized macros not working for contacts

Andreas Ericsson — 2012-05-03T14:46:06

Are environment macros are turned on and large_installation_tweaks
turned off? Otherwise, it's not supposed to work.


When someone sends in a patch, or when Eric and Ethan finds time to
work on this. Environment macros are nifty to use in small installs
but the performance penalty associated with them along with the weird
errors that occur in large installations with many large groups of
objects make them unsuitable for real-world usage, so I prefer to
deter people from using them rather than fixing any problems with them
(since the most severe problems just aren't fixable).

Customized macros not working for contacts

Deborah Martin — 2012-05-03T10:42:53

Hi,

I'm running nagios 3.3.1 on sles 10.

I've defined a contact which uses a customized macro called _EMAILCONTACTID

However, when I try to reference this macro, from a script defined to run as part of a service check,  as an environment variable $NAGIOS__CONTACTEMAILCONTACTID, is blank

If I define a customized macro in the service definition, I can see the contents of $NAGIOS_SERVICEEMAILCONTACTID.

I've followed the docs and it does state that you can define a customized macro in a contact definition, so this to me looks like a bug in 3.3.1

When will a fix be available ?

Regards,
Deborah


This e-mail and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed.  If you are not the intended recipient, please delete this e-mail immediately.  Any unauthorised distribution or copying is strictly prohibited.

Whilst Kognitio endeavours to prevent the transmission of viruses via e-mail, we cannot guarantee that a

Re: Log USERNAME when DISABLING/ENABLINGchecks/notifications...

Andreas Ericsson — 2012-04-06T14:35:05

It's in testing at our lab as we speak. The code is currently only
available from git://github.com/ageric/nagios.git where it'll remain
until I've ironed out the following remaining issues:

* epoll()-based I/O broker doesn't work with some kernel+glibc versions.
I have no idea why, and it's possible to fallback to using poll() or
select() instead, but it's pretty unclear how one chooses which one to
use, so that'll have to be made better.

* API docs have to be generated and published somewhere so other module
authors can figure out how to use it.

* Tests have to be done to make sure external commands retain their
previous performance. I have a hunch that it might be affected by
having to share the polling space with the workers.

* Handling has to be added for when pollers go AWOL. It's sort of
amazing that it hasn't happened by accident yet, but I guess it will
in real-world scenarios, and we should handle it gracefully.



We'll most likely solve this with a subscriber service where you can
get raw log

Re: Log USERNAME whenDISABLING/ENABLINGchecks/notifications...

Daniel Wittenberg — 2012-04-06T05:48:14

We have issues with this as well. Maybe an option to log in micro epoch? I also wriote an event broker that opens a tcp connection to splunk and directly pipes all logging over, which allowed us to get around that issue but doing it in core would be nice.
Dan

From: Deepak Kosaraju [mailto:deepak.kosaraju1< at >gmail.com]
Sent: Thursday, April 05, 2012 06:22 PM
To: Andreas Ericsson <ae< at >op5.se>
Cc: Nagios Developers List <nagios-devel< at >lists.sourceforge.net>
Subject: Re: [Nagios-devel] Log USERNAME when DISABLING/ENABLING checks/notifications...

Andreas
Thanks for the prompt reply, So when can we expect Nagios 4, where can I track the beta release schedules for Nagios.

When you are planning to change logging version it would be nice to look at the time [filed].

Here is the scenario: Between nagios and splunk,
As splunk indexer sees the event from forwarder, the indexer records the time based on the event time and converts that to HRT format, and as nagios log doesn't have anything to say 2 events