Sunday, May 22nd 2011, 8:26pm UTC+2

You are not logged in.

Icinga Multi-Site Question (not Site-to-Site)

EagleDelta1

Beginner

Posts: 5

Birthday: Apr 2nd 1984 (27)

Gender: male

Location: Omaha, NE, USA

Occupation: IT Infrastructure Coordinator

Number of Nagios server: 2

Nagios Versions: 3.0

Icinga-Version(en): Icinga 1.4.0

Distributed monitoring: Nein

Redundant monitoring: Ja

Number of hosts: 24

Number of services: 210

OS: Ubuntu Server 10.04.2

Plugin Versions: 1.4.15

1

Thursday, May 19th 2011, 12:49am

Icinga Multi-Site Question (not Site-to-Site)

Ok, so after considering the possibility of using a Site-to-Site setup with a Centralized database for Icinga monitoring, I've decided that due to the possibility of one site losing connection to the others (happened recently when Verizon decided to screw with the lines in our office building without notifying us) using a single, centralized DB for monitoring would be a bad idea.

As such I'm planning on moving forward with a full Icinga server at each site, each with its own Database, as well as a cold Icinga backup system for emergencies. From here I need to know is if, by using different instance names for the Icinga Database at each site, I would be able to "export" the data from the four independent databases to our Data Warehouse for centralized reporting for our monthly SLA uptime reports.

Here is an example of what I'm trying to do:

Diagram of the 4 sites and their connection



Diagram of what I'm hoping to do:



Any suggestions or warnings?

Thanks!

David Hollinger III

bernd_erk

Trainee

Posts: 124

Gender: male

Location: Nürnberg

Number of Nagios server: -

Nagios Versions: -

Icinga-Version(en): 1.3

Distributed monitoring: Ja

Redundant monitoring: Ja

Number of hosts: -

Number of services: -

OS: Debian

Plugin Versions: -

2

Thursday, May 19th 2011, 9:55am

Hi,

i see no problem with that.

Another way you can do it, is the usage of passive events and transfer over the monitoring-information using perf-data-output. You can store them locally if the connection is down and can store everything in one database later on

Regards

Bernd
NETWAYS GmbH http://www.netways.de
NETWAYS Blog http://blog.netways.de

EagleDelta1

Beginner

Posts: 5

Birthday: Apr 2nd 1984 (27)

Gender: male

Location: Omaha, NE, USA

Occupation: IT Infrastructure Coordinator

Number of Nagios server: 2

Nagios Versions: 3.0

Icinga-Version(en): Icinga 1.4.0

Distributed monitoring: Nein

Redundant monitoring: Ja

Number of hosts: 24

Number of services: 210

OS: Ubuntu Server 10.04.2

Plugin Versions: 1.4.15

3

Friday, May 20th 2011, 12:12am

I actually like that idea better - less resources to try and manage.

Just to clarify - would the remote Icinga hosts need to be checking everything passively or is that central Icinga server accepts all the remote checks via the check_nsca plugin (similar to the way dist monitoring is setup as described in the docs) or am I completely off base?

Also, I have very little experience so far in taking advantage of perf-data (outside of PNP4Nagios), so how would I go about using perf-data-output to send the information to a central database?

dnsmichi

Master

Posts: 2,102

Birthday: May 30th 1983 (27)

Gender: male

Location: Wien

Occupation: DNS & Monitoring Developer

Number of Nagios server: 3x Icinga prod, 2x test/dev

Nagios Versions: s/nagios/icinga/

Icinga-Version(en): 1.4.0 / GIT

Distributed monitoring: Ja

Redundant monitoring: Ja

Number of hosts: 1000+

Number of services: 15000+

OS: RHEL 5.6 x64

Plugin Versions: 1.4.15

IDO-Version: 1.4.0 / GIT MySQL/Postgresql/Oracle

Other Addons: PNP 0.6.11, check_mk GIT

4

Friday, May 20th 2011, 8:22am

better to be explained within the docs http://docs.icinga.org/latest/en/distributed.html
Icinga 1.4 MySQL,PGSQL,Oracle and IPv6

Demos

Docs http://docs.icinga.org
Wiki https://wiki.icinga.org

Dev Tracker https://dev.icinga.org

* egrep -v "^#|^$" configfile

Join us online!
irc.freenode.net #icinga-devel

Using Icinga? Tell us!

bernd_erk

Trainee

Posts: 124

Gender: male

Location: Nürnberg

Number of Nagios server: -

Nagios Versions: -

Icinga-Version(en): 1.3

Distributed monitoring: Ja

Redundant monitoring: Ja

Number of hosts: -

Number of services: -

OS: Debian

Plugin Versions: -

5

Friday, May 20th 2011, 9:23am

agree to dnsmichi :-)
NETWAYS GmbH http://www.netways.de
NETWAYS Blog http://blog.netways.de

EagleDelta1

Beginner

Posts: 5

Birthday: Apr 2nd 1984 (27)

Gender: male

Location: Omaha, NE, USA

Occupation: IT Infrastructure Coordinator

Number of Nagios server: 2

Nagios Versions: 3.0

Icinga-Version(en): Icinga 1.4.0

Distributed monitoring: Nein

Redundant monitoring: Ja

Number of hosts: 24

Number of services: 210

OS: Ubuntu Server 10.04.2

Plugin Versions: 1.4.15

6

Friday, May 20th 2011, 5:07pm

Awesome, just wanted to make sure I was looking in the right place.

Thanks guys!

MiCkEy2002

Administrator

Posts: 4,175

Birthday: Feb 29th 1976 (35)

Gender: male

Location: Roffhausen

Occupation: Systemadministrator

Number of Nagios server: 9

Hobbies: Bungee Springen, lesen....

Nagios Versions: 2.5 / 3.03

Distributed monitoring: Ja

Redundant monitoring: Ja

Number of hosts: ~1700

Number of services: ~18000

OS: SuSE SLES 9

Plugin Versions: 1.4.3

NagVis Version: 1.2.2

NDO Version: 1.3.1/1.4b7

Perfparse Version: 0.106.1

Other Addons: NagTrap 1.3/PNP

7

Friday, May 20th 2011, 6:49pm

Hi,

i use the distribute monitoniring. The crentral instance write the data in a central database. When one node is down and no passive checks was send to the central instance, then the central instance makes the checks for the node. It works very good :D

Bye
Michael
Projekte:
Die OpenSource Community hilft
NagTrap-Developer
Icinga-Developer
NagVis

Similar threads