Business IT is increasingly heterogeneous, with elements in house (legacy) and elements hosted by third party providers (Cloud). Monitoring of all the different elements is essential, as we have mentioned several times in earlier blog articles.
Let's see, using a real case study how to set up monitoring of IaaS and PaaS infrastructures hosted at Azure, at the same time as retaining internal IT elements.
Our example is a car park manager, with a connected system to indicate to the users of the car park the available spaces. In this way they can maximize the filling of the parking lot, while providing users with a stress-free experience and reducing the time spent searching for a space.
Our car park manager has a system based on 3 discrete elements :
- Azure PaaS components, with :
- an IoT Hub for real-time sensor status feedback
- an Azure BDD, a database to store real-time data
- Multiple WebApps to expose APIs and a web interface for the manager
- a "notification hub" to send notifications to the manager as required
- Azure IaaS components to host:
- a Windows machine with a SQL Server ("BI" database)
- a Windows machine with PowerBI for the management and the production of various reports
- On-premise components for:
- Information retrieval
- Manage the display of light panels
- Manage the local network
Implementation of monitoring with ServiceNav :
Step 1: Supervision of the elements
For PaaS bricks: Supervision via the MS-Azure-PaaS-Metrics service model and feedback of the metrics required for supervision and comparison with thresholds.
For IaaS bricks: Two options were possible: Setting up a ServiceNavBox on the Azure platform as a VM in the same VLAN as the machines to be supervised or supervised by the APIs with the Host Alive by Azure API equipment models and associated service models from the ServiceNavBox set up in its Internal IT. We finally opted for a ServiceNav Box as a VM in the Azure cloud.
For internal bricks: "classic" supervision with the installation of a ServiceNav Box as close as possible to the equipment in the local information system. The ServiceNav Box is connected to the ServiceNav SaaS platform. It queries the equipment (server, windows services, network equipment, light panels, etc.) via SNMP and WMI.
Second step: Setting up a MAP
From a PowerPoint diagram of the client's infrastructure, and by placing the different elements on a MAP, we were able to build a very visual representation of the infrastructure giving its state in real time and the important metrics.
Step 3: Setting up a service weather forecast
By aggregating all checkpoints, it is then possible to build a service weather indicating the complete status of the parking space management system. It is thus possible to calculate the SLA on the complete availability of the service.
Thanks to ServiceNav and its integrated service and equipment models, it is possible to supervise a mixed infrastructure: Cloud Azure and internal "Legacy" IT.
From technical supervision to functional operation, real-time vision of the complete infrastructure and calculation of SLAs, ServiceNav provides all the teams involved in the project at the customer's site.