Here you can find all videos and slides of the OSMC 2013:
Holger Koch | 10 Jahre Monitoring mit Open Source Software bei der DB Systel (DE)
In this talk you will get an overview of the beginnings . I will amplify the used technologies, the architecture as well as the cultural problems we had in the beginnings.
The second part of the presentation will show why and how DB Systel managed to escape from an update pitfall and how they accomplished the migration of an Icinga-based monitoring platform.
Reiko Streng | Monitoring-Landschaft auf Basis von OMD (DE)
it@m manages as central IT service provider the entire heterogenous landscape of the Landeshauptstadt München with about 1500 hosts and 15000 services. To meet the growing requirements and needs, a total refactoring and restructuring of the existing monitoring was made.
This talk describes the monitoring landscape based on OMD, swapping out by use of Mod-Gearman the different tasks (PNP4Nagios, dedicated plugin workers,…) to particular workers to ensure fluent scalability. The required configuration is automatically generated several times per day from different sources (CMDB and others) and, by using GIT, is provided audit-proof to the Nagios core.
In this project, the Landeshauptstadt is committed to open source products.
Thomas Lehmann | Monitoring als Quelle der Wahrheit im Wellendeployment einer dynamischen Webseite (DE)
ImmobilienScout24 runs nearly 2000 server instances with 160 applications, enhanced by 100 developers and maintained by 30 administrators, entailing system changes at hourly intervals. We will cover topics like DevOps mindset, automation, continuous delivery, blasting persistency and the permanent implementation of all changes.
The central component at ImmobilienScout24 to store data and to analyze is Graphite, containing all monitoring information. This component is actively analyzed by Icinga and compared with thresholds. The aim is the client-based configuration of Icinga Alarming by rolling out the corresponding values on the client.
The current aim is to enhance our monitoring in a way we can achieve the core of wave deployment. Icinga Alarming monitors instances during the deployment, controls the wave rollout process through our open source monitoring solution YADT and informs the application manager about errors. Here, we frequently reach the concept and performance limits of existing open source solutions. That is why we, in cooperation with the project Maintainer Upstream, develop them further and make them available to the open source community.
In this talk you will learn how to build up a reliable and comprehensive monitoring and alarming system in such an agile environment.
Johannes Egger | Enterprise Platforms Monitoring at s IT Solutions AT (DE)
The enterprise platforms department of the IT service provider IT Solutions AT is in charge for a large number of system components of UNIX, Windows, applications, databases, virtualization, storage and hardware management. An Icinga environment is used to monitor these system components. Currently about 46000 host and service checks are carried out in one environment, designed for 65000 – 80000 checks and counting.
In this talk the migration from Tivoli Endpoints to Icinga with focus on the issues related to a large Icinga installation will be described. Main features are the (partial) automation of the configuration generation from the CMDB, the assignment of hosts to the different host groups based on various criteria, the automation of the agents and plugins deployment as well as the interface to the superordinate umbrella monitoring of the group.
Bernd Strößenreuther / Sascha Brechmann | Puppet konfiguriert Nagios automatisch (DE)
Part 1: Basics
Every time a new server is set up, it must be added to the monitoring (CPU, RAM, Disk…).
And every time a new service (e.g. Apache webserver) is installed on a machine, it has to be monitored (e.g. with check_http).
Those who use Puppet to configure machines and services can extend the Puppet modules in a way the monitoring for example is set up (via Exported Resources) automatically on the corresponding machine. It is not only time saving, it also ensures that every system is integrated into the monitoring.
This talk will explain the application of Puppet and Nagios (classical version) by specific examples.
Part 2: Puppet / OMD / Nagios / Check_MK in practical use
This presentation will show everything that can be done with Puppet and how. Topics will be for example: The installation of OMD / Check_MK server. The installation of the Check_MK Agent on the client. The automatic integration incl. inventory of the clients on the check_MK Server. Setting of Check_MK_Agent tags per facts and within Puppet modules as well as when installing Check_MK_Agent. The integration of Mod_Gearman for laod balancing and to break through firewall limits. Establishing an automatic documentation in the OMD Dokuwiki based on Puppet facts of all virtual machines. The use of Puppet tags to generate additional active checks. (Check_HTTP; SMTP; etc.)
Jeff Gehlbach | What´s New in OpenNMS (ENG)
OpenNMS is the world's first enterprise-grade network management application platform developed under the open source model. Learn what's new in OpenNMS 1.12 and what's coming in OpenNMS 2.0, such as VMWare support, topology and geographic mapping as well as highly scalable data collection.
Rihards Olups | Zabbix: A Practical Demo (ENG)
While having a high level overview of how a system works is useful, nothing can explain better how system works than seeing it in action.
This talk will be like a practical quickstart, demonstrating actual basic configuration of Zabbix starting with data gathering and ending with receiving an alert.
Icinga - Team | Current State of Icinga (DE)
Much has happened since the first preview on the OSMC 2012 and Icinga has taken shape. Beside enhanced performance and flexible architecture and centralized and decentralized cluster, Icinga2 is above all characterized by simplified configurations. Also Icinga1 as well as the two add-ons and web interfaces went through changes. This presentation will give an overview of the current status and new developments and will introduce the latest Icinga version in live demos.
Alan Robertson | Monitoring and Discovery without Limit (ENG)
The Assimilation Project provides integrated IT discovery and monitoring. Discovery is continuous and has zero-network-footprint.
Monitoring is extremely scalable due to a radically distributed architecture. Discovery informs monitoring - simplifying configuration and maintenance. Discovery finds systems you’ve forgotten about, and services you aren’t monitoring and dependencies. About 30% of all outside security breaches come through forgotten systems *- *critical for security.
Luca Deri | Monitoring network traffic using ntopng (ENG)
ntop is open-source network monitoring project aimed at network traffic monitoring. With the advent of modern computing architectures and high-speed networks many components had to be redesigned to address the new challenging issues. Monitoring 10G networks require high-speed packet capture/transmission and efficient monitoring applications. Over the past 10 years ntop has created several open-source tools, kernel modules, flow-based applications for making high-speed traffic monitoring feasible on commodity hardware. This talk will give an overview of the various ntop monitoring components including the latest version of ntop named ntopng. It will also highlight some use cases where these tools have been used successfully.
Andreas Ericsson | The future of Nagios (ENG)
Oliver Hankeln | openTSDB - metrics for a distributed world (ENG)
At gutefrage.net we like to measure everything - like the saying goes:
"what gets measured gets managed", we measure everything from server load to content generation to adsense income.
To store more than 100M data points for more than 2000 metrics per day, you need the right tool for the job. For us this is openTSDB.
In this talk I will give an overview over the architecture of openTSDB, compare it to other systems in the field, show how easy it is to gather new metrics and give advice on running the system and choosing metrics.
Anders Håål | Why dynamic and adaptive thresholds matters (ENG)
The presentation will cover why dynamic and adaptive thresholds is key to modern IT and business monitoring.
The presentation will go through a number of use cases where complex threshold logics are applicable:
- Threshold based on historical data
- Thresholds based on aggregated data from multiple services performance data and/or state
- Adaptive thresholds when the monitored service is related to other services
- Thresholds based on the time of the day and other calender profiles
- Using mathematical functions like average, derivative, sum, min, max, etc to calculate thresholds based on series of data
In the presentation we will show how this has been achieved using the Bischeck open source framework, www.bischeck.org, integrated with Nagios and other surveillance systems
Fernando Hönig | Distributed Monitoring and Cloud Scaling for Web Apps (ENG)
The first part will cover how we have integrated Nagios and WebInject in a distributed monitoring infrastructure in AWS EC2 using event handlers for making extra checks from different locations setting a threshold for each location and fire an status for each situation. We also call different APIs from other monitoring tools like Gomez Networks and Pingdom and use that as feeds. For notifications we integrated CallWithUS (For VoIP calls) and Clickatell for SMS alerts.
The second part of the presentation will show in deep how to enable/disable hosts, services from devices that are created automatically with Amazon EC2 auto-scaling in a cloud infrastructure.
Also will show how important are the hostgroups and the implementation of those in a diverse kind of customers. After that will be how we integrate Nagvis to Nagios using mklive_status broker in order to make awesome maps and even more maps automatically. This part also will show how easy is to join different nagios systems in only one map and have the whole picture of our systems in one shoot.
Georg Kostner | End User Experience Monitoring with Al'exa (ENG)
Al’exa is a Python module that provides capabilities to automate and test graphical user interfaces by using computer vision to verify the availability and performance of user application which are deployed for example over Citrix and Windows Terminal Server. Al’exa is built on the OpenCV vision library to recognize objects on template based patterns, to identify UI elements (labels, input fields, buttons, links, trees, ...) and to interact with them like a normal user does. An IDE based on Ninja Python enables the creation of these test cases. The execution of those test cases with Al’ax is already nsclient++ compliant so that Nagios, Icinga, Shinken can schedule them and get the result (OK, WARNING, CRITICAL) and performance data back.
Oliver Tatzmann | Zielgruppenorientiertes IT Servicemonitoring (DE)
The monitoring information shall not only be available for administrators but also for people with less technical knowledge (service owner, management, end-user). The aim is to gather all requirements of the different target groups, to sum them up, to realize them and to make them available for all target groups, beginning with the monitoring of a single component to the consideration of complex systems.
This talk will focus on the practical experiences of the requirement gathering and the realization with Nagios/Icinga and standard add-ons based on a global available service.
Bernd Erk | Log- und Eventmanagement (DE)
The management of passive events including their correlation is not one of the great strength of Nagios and Icinga. Of course it is possible to monitor and analyze different sources employing diverse plugins and add-ons, however, a rule-based processing and long-term storage is usually difficult.
There are diverse open source monitoring solutions available to ensure an end-to-end traceable chain, beginning with emergence up to storage. Tools like Logstash, EDBC and statsd has become increasingly popular over the past months but often are configured and deployed independently. The big challenge is to connect the existing tools in order to make optimal use of their individual strengths.
After a brief introduction of the used tools, this presentation will provide an insight into the implementation of described components and their connection with Nagios and Icinga. Starting with the installation and configuration and ending with filtering and correlation rules.
Sven Nierlein | Thruk - Monitoring GUI (DE)
Thruk is a WebGUI which displays clearly several Nagios, Icinga and Shinken instances using the Livestatus addon. It was designed to replace the Nagios CGIs by a performant alternative and provides beside the standard features also add-ons for reporting, dashboards, configuration and mobile support as well as little extras to simplify the daily work with Nagios.
This presentation will give an overview of Thruk, show tips and tricks for the installation und application as well as existing add-ons and plugins.
Michael Medin | Making monitoring simple? (ENG)
Showcasing the new simplified NSClient++ monitoring agent.
With the up-coming release of 0.4.2 NSClient++ will finally have a new re-written check subsystem which means not only will things work on "modern windows" they will do so with a flare. I will show how the new simplified syntax as well as real-time monitoring works through out all commands in NSClient++. In addition to this there will be updates on the Linux front as well as agentless monitoring...
Pall Sigurdson | Adagios- Web configuration done right (ENG)
Adagios is a nagios/icinga add-on that allows you to configure nagios in-place without a database backend. It offers Web API, full-featured configuration, and has a built-in status interface that interfaces natively with PNP and mk-livestatus.
This presentation will focus on the challenges of maintaining a large monitoring environment and how the adagios and its underlying pynag toolkit can help running a modern monitoring system.
Birger Schmidt | Flapjack - monitoring notification system (ENG)
Flapjack (http://flapjack-project.com/ ) sits on top of existing monitoring systems and allows to set them up to a highly scalable overall system. Flapjack processes monitoring results into notifications. Thereby the monitoring and notification process are being degenerated.
For instance a combination of different Nagios, Icinga and Sensu instances can execute checks and transmits the results to Flapjack for analyses. Flapjack generates notifications which can be delivered via PagerDuty, XMPP, email and SMS.
The functionality of Flapjack is available for other components due to the API.
This presentation will give an insight into the motivation, history, backgrounds and the structure of the software.
Gerhard Laußer | coshsh - Generationen von Konfigurationen (DE)
In 2001 the framework coshsh has been created to automatically generate Nagios configuration files. As data sources random CMDBs and other inventory lists can be used. Out of the encountered objects (hosts and installed applications), coshsh creates with the help of rules and template files the final configurations. It is important to mention that there is no monitoring know-how required for the input part, hence Windows administrators, DBAs, etc. are able to register easily their systems. The special know-how and which services emerge lately are determined by the above mentioned rules and templates.
In this talk monitoring admins will explain the deployment of coshsh and will show the different fields of application.
Werner Fischer | TKmon - easy web-based configuration of Icinga (ENG)
Icinga and Nagios are very poplar throughout nearly all Linux sysadmins around the world. However, for sysadmins of other operating systems it was in the past difficult to install a Linux system and to set up Icinga with a reasonable configuration to monitor the sysadmin's network.
To overcome this burden, NETWAYS and Thomas Krenn have developed an easy-to-use web frontend, which allows every sysadmin to configure and use Icinga straightforward.
TKmon's main features include showing the current service status, and setting up hosts, services and contacts. Extensible service catalogs allow the definition of services, so that they support optional parameters to be entered via TKmon's web interface.
Additional features of TKmon are: network configuration, mail configuration, security configuration (to disable/enable direct access to Icinga or SSH), and a backup/restore function for the whole configuration.
TKmon has been released under the GPL at this year's LinuxTag back in May. Since then, ready-to-use TKmon packages for Ubuntu 12.04 LTS are available at Thomas Krenn's repository archive.thomas-krenn.com. A ISO image to install a Ubuntu 12.04 based TKmon appliance is also planned and should be available at OSMC.
So now there is no reason anymore, why Windows sysadmins should not use Icinga ;-)
Further information about TKmon can be found at: www.thomas-krenn.com/en/wiki/TKmon
Martin Fürstenau | check_vmware_esx.pl - ein Rewrite von check_vmware_api.pl (DE)
Why is a rewrite of check_vmware_api.pl useful?
The monitoring of vmware in a Nagios/Icinga landscape is quite a black hole. The most popular plugin check_vmware_api.pl (former check_esx3.pl) has already driven many admins to desperation. The advantage is not always obvious; some things are not documented and some options don´t seem to make sense. The code structure with its 4600 lines does not simplify modifications and extensions. The reason why we rewrote the plugin was the storage monitoring. With dozens LUNs, this plugin was completely useless. Furthermore, we noticed some bugs and saw more potential. As the realization of suggested modifications is rather slow, we decided to create a fork and rewrite.
This presentation will show the current state, explain which things we dropped and why and where performance counter can be misinterpreted and of course when this plugin will be available.