ProfessionalMonitoring automation
Serverless metrics reporting across multiple Zabbix servers
A Python Azure Function that collects key metrics from several Zabbix servers, stores history in Azure Storage, generates Excel reports and delivers them to Microsoft Teams — replacing manual reporting.
- Role
- Lead developer
- Context
- ALIANDO · Managed Services area
- Period
- 2025
01Overview
Monitoring data is only useful if it reaches the people who need it. I led the development of a fully automated pipeline that pulls key metrics from multiple Zabbix servers, keeps their history in Azure Storage, builds Excel reports and posts them to Microsoft Teams — with no one assembling spreadsheets by hand.
02Context
In ALIANDO’s Managed Services area, client environments are monitored with Zabbix, often across several Zabbix servers. Periodic reporting on key metrics was part of the service.
03Problem statement
Gathering metrics from several Zabbix servers, consolidating them and producing reports by hand is slow, repetitive and error-prone, and leaves no centralised history to compare periods.
04My responsibilities
- Led the design and development of the data-collection system.
- Implemented the collection and reporting logic in Python on Azure Functions.
- Integrated multiple Zabbix servers as data sources.
- Automated report delivery to Microsoft Teams.
05Architecture overview
A serverless Azure Function written in Python queries each Zabbix server through its API, writes the collected metrics to Azure Blob Storage as a centralised history, generates an Excel report and delivers it to a Microsoft Teams channel. The Function and its storage are deployed with Terraform.
- Observability
- Azure service
- Data & storage
- External / reference
- Code & configuration
Text description of the diagram
Several Zabbix servers are queried through the Zabbix API by a Python Azure Function. The function stores historical metrics in Azure Blob Storage, generates an Excel report and delivers it to Microsoft Teams. Terraform deploys the function and storage.
Flows
- Zabbix server A → Azure Function (Zabbix API)
- Zabbix server B → Azure Function
- Zabbix server N → Azure Function
- Azure Function → Blob Storage (store)
- Azure Function → Excel report (generate)
- Excel report → Microsoft Teams (deliver)
- Terraform → Azure Function
Components
Zabbix server A · Zabbix server B · Zabbix server N · Azure Function · Blob Storage · Excel report · Microsoft Teams · Terraform
06Technology stack
- Compute
- Azure Functions (serverless)
- Python
- Data
- Azure Blob Storage
- Excel
- Integrations
- Zabbix API
- Microsoft Teams
- Deployment
- Terraform
07Implementation approach
- Serverless execution on Azure Functions — no servers to patch for a periodic workload.
- One collector that iterates over multiple Zabbix servers, so adding a source is configuration, not code.
- Historical data centralised in Blob Storage, making period-over-period comparison possible.
- Report generation and delivery automated end to end, with Teams as the delivery channel the team already used.
- Infrastructure for the solution declared in Terraform.
08Challenges & engineering decisions
Challenge
Data spread across several independent Zabbix servers.
Decision
Centralise collection in a single function and persist results in one storage location, giving a single source of truth for reporting.
Challenge
Reports are only valuable if they actually reach their audience.
Decision
Deliver directly into Microsoft Teams instead of leaving files in storage for someone to fetch.
09Security considerations
- Zabbix API credentials and the Teams delivery endpoint are secrets — they belong in application configuration, never in source code.
- The function needs read access to Zabbix and write access to its own storage; nothing broader.
10Operational considerations
- Fully automated — collection, storage, report generation and delivery run without manual steps.
- Serverless compute means no infrastructure to maintain for a workload that runs periodically.
11Results & impact
- Manual multi-server reporting replaced by an automated pipeline.
- Reduced operational costs and improved reporting efficiency.
- Centralised metric history available for analysis.
12Lessons learned
- Automate the whole path to the reader, not just the data collection.
- Serverless is a natural fit for periodic glue workloads between platforms.