Checkmk monitors retail environments at Edeka
Who is Edeka Minden-Hannover?
With a turnover of 10.6 billion Euros in 2020, 32,000 employees and around 1,600 stores, Edeka Minden-Hannover is the largest of seven Edeka regional cooperatives in Germany. Its territory reaches across all of northern Germany from the Dutch to the Polish border. It covers a part of Eastern Westphalia, most of Lower Saxony, and the federal states of Bremen, Sachsen-Anhalt, Berlin and Brandenburg.
The speed and target achievement, all the way from planning to implementation, surprised us. The collaboration with tribe29 was professional and determined from the very beginning.
Edeka Minden-Hannover has been using Checkmk since 2012. Before Checkmk, the IT team used Nagios as monitoring solution, which was no longer sufficiently scalable and was not able to meet Edekas’ requirements anymore.
Checkmk was originally intended to be used as a front-end for managing the Nagios environment, but after a short time, the IT team decided to completely replace Nagios with Checkmk. The migration of the 1,200 Nagios servers went smoothly.
The challenges were not only the large scale of the existing Nagios monitoring, but also the special requirements of the retail sector.
Edge monitoring of retail environments
Edeka Minden-Hannover was looking for a cost-efficient, flexible and modern solution to monitor its stores and central IT systems. The organization originally selected the Nagios open source software. The IT team installed Nagios manually and extended it bit by bit with add-ons such as NSCA or NagVis. However, the performance of Nagios soon reached its limits.
A first problem was the heavy network load due to the many details the monitoring has to cover. Also, even with graphical configuration tools such as NConf, the workload for the IT team just for managing the monitoring environment became unbearable. In time, the Nagios server also turned out to be a performance bottleneck. Thus, the IT team had to start looking for a Nagios alternative. Edeka Minden-Hannover discovered Checkmk and with it was able to replace its Nagios environment easily. Thanks to Checkmk, the IT team can continue using existing Nagios plug-ins after some minor adjustments. Because Checkmk runs on any major Linux distribution, Edeka Minden-Hannover was also able to install Checkmk on its existing SUSE Linux Enterprise server infrastructure.
Edeka Minden-Hannover’s monitoring requirements are demanding: The IT team wants to gather the monitoring data at a central site to be able to gain an overview of all their IT environments. The monitoring should automatically generate a map, based on GPS coordinates, as a central status monitor. In critical situations, it is also important to send automatic notifications, for example, if fewer than a specified percentage of cash registers are ready to operate in a store.
The replacement of more than 1,200 independent Nagios servers is certainly not an everyday event. Generally, the management of such environments is more than a full-time job. But this is not the only impressive number within this set-up: As of today each of those 1,200 servers monitors an average of 25 hosts and 250 services. This adds up to 300,000 services, which Checkmk checks once every minute. All of this information is then aggregated in one central site.
In most markets, there are no IT specialists available, so another requirement is that employees without IT experience should also be able to understand and use on-site monitoring. At the same time, the monitoring should automatically add new systems into the monitoring. In addition, monitoring in the stores must also continue to work, even if a store is disconnected from the rest of the network or the connection is unstable. The only way to ensure this is to monitor local systems through a local monitoring site.
At the same time, Edeka Minden-Hannover wants the monitoring to aggregate data centrally. Checkmk brings its own agents that do not need to be configured on the monitored systems. This made the rollout very easy, even with a large number of systems. Thanks to its inventory function, Checkmk automatically detects which aspects of a system can be monitored. Checkmk regularly scans the local network of the respective store for new components. To do this, the IT team relies on the standard nmap tool. As soon as nmap finds a new system, Checkmk uses the automatic inventory function to determine which of the systems’ services need to be monitored.
The IT team configures the threshold values via flexible rules on the central monitoring site and applies the monitoring rules globally across all stores. Checkmk thus minimizes the effort required to manage the monitoring in the stores. The rule set consists of 26 different rules. The aggregated overall status from each market can then be evaluated in the central Checkmk site. This aggregation is realized with the help of the Checkmk Business Intelligence (BI).
Checkmk uses its JSON-based web service for the implementation. This ensures that the information in the central site is up to date and reflects the current state of the store IT. A dashboard based on the data from the central site displays the overall status on two 55-inch monitors. On this dashboard, there are individual views (dashlets) that show, among other things, the stores with connection problems and host or service problems in their own lists. As part of the project, NagVis was extended to include geomap functionality. NagVis creates a map using the GPS coordinates of all locations and freely available map material from OpenStreetMap and positions the locations on this map.
Edeka Minden-Hannover successfully replaced Nagios with Checkmk in 2012. The monitoring system collects detailed information for a wide range of systems from all of its stores. The central Checkmk site summarizes this information before visualizing and evaluating it.
Successful rollout together with tribe29
During the initial installation of Checkmk, tribe29 also developed some new Checkmk checks (such as the monitoring of Bintec routers), which were subsequently incorporated into the official version of Checkmk. Following a few weeks of planning, Edeka Minden-Hannover and tribe29 carried out this major rollout to all stores in a four-day on-site process.
Within six hours, Edeka Minden-Hannover had installed a total of 1,200 monitoring systems. That means three new systems per minute! Following the installation, Edeka Minden-Hannover connected all of the systems in the stores to the central site. At the same time, the IT team put the geomap with the information from the stores into operation.
With little effort, and in partnership, Edeka Minden-Hannover and tribe29 developed and successfully implemented a suitable monitoring solution. The entire project was developed based on license-free software. At the same time, tribe29 provided expertise and further expanded the customer's monitoring know-how.
With Checkmk, Edeka Minden-Hannover now has a very scalable and precise monitoring system in place. The company is thus optimally positioned for current and any future monitoring challenges. The switch from Nagios to Checkmk had already paid off for the company after a short period of time, as the new system has significantly reduced the workload for the IT team.