adr: new adr - Prometheus operator

Signed-off-by: Viktor Login <[email protected]>

ops/docs/ADR/README.md

@@ -40,6 +40,10 @@ for an information team about cases for updated ADR.
 - [ADR-0027](./decisions/0027-local-kubernetes-development-tools.md) - Local Kubernetes Development Tools
 - [ADR-0031](./decisions/0031-k8s-network.md) - k8s network
 
+#### Observability
+
+- [ADR-0011](decisions/0011-prometheus-operator.md) - Prometheus operator
+
 #### Helm
 
 - [ADR-033](./decisions/0033-helm-use-values-schemas-json.md) - Use `values.schema.json`

ops/docs/ADR/decisions/0011-prometheus-operator.md

@@ -0,0 +1,61 @@
+# 11. Prometheus Operator
+
+Date: 2024-10-25
+
+## Status
+
+Accepted
+
+## Context
+
+As our Kubernetes infrastructure scales, we require a robust and efficient monitoring solution to handle the increasing complexity and volume of metrics. Historically, we've used Prometheus for monitoring due to its reliability and widespread adoption. The introduction of the Prometheus Operator has simplified the deployment and management of Prometheus instances in Kubernetes environments.
+
+However, we also considered **VictoriaMetrics** as an alternative. VictoriaMetrics is known for its high performance, efficient storage, and scalability, particularly in large-scale environments. It offers compatibility with Prometheus but claims better resource utilization and faster query performance.
+
+Key considerations influencing our decision include:
+
+- **Scalability and Performance**: Ability to handle high ingestion rates and large volumes of metrics data.
+- **Ease of Deployment and Management**: Simplifying operations within our Kubernetes clusters.
+- **Community and Ecosystem Support**: Availability of support, documentation, and integrations with existing tools.
+- **Future-Proofing**: Incorporating advancements from recent developments like Prometheus 3 Beta.
+
+## Decision
+
+We have decided to adopt the **Prometheus Operator** for our Kubernetes monitoring needs.
+
+**Rationale**:
+
+- **Kubernetes-Native Management**: The Prometheus Operator leverages Kubernetes Custom Resource Definitions (CRDs) to manage Prometheus instances, making it a seamless fit for our Kubernetes-centric infrastructure.
+- **Maturity and Community Support**: Prometheus has a large and active community, ensuring ongoing development, support, and a wealth of resources for troubleshooting.
+- **Ecosystem Compatibility**: Prometheus integrates well with our existing exporters, dashboards, and alerting systems, minimizing the need for extensive reconfiguration.
+- **Future Enhancements**: The upcoming features in [Prometheus 3 Beta](https://www.prometheus.io/blog/2024/09/11/prometheus-3-beta) promise improved performance and scalability, aligning with our long-term needs.
+- **Operational Simplicity**: Using the Prometheus Operator reduces operational overhead by automating routine tasks like configuration reloads, version upgrades, and horizontal scaling.
+
+While **VictoriaMetrics** offers impressive performance and storage efficiency, the transition would require significant changes to our current setup and retraining of our team. The benefits do not outweigh the costs and risks associated with migrating to a new system at this time.
+
+## Consequences
+
+**Benefits**:
+
+- **Simplified Deployment and Management**: The Prometheus Operator automates the setup and maintenance of Prometheus instances, reducing manual interventions and potential for human error.
+- **Enhanced Scalability**: We can easily scale our monitoring infrastructure to handle increased workloads, thanks to the operator's support for horizontal scaling and sharding.
+- **Consistent Ecosystem**: Maintaining our existing Prometheus-based stack ensures compatibility with current tools and processes, facilitating smoother operations.
+- **Leverage New Features**: We can take advantage of the enhancements in Prometheus 3 Beta, such as improved storage efficiency and query performance.
+
+**Drawbacks**:
+
+- **Resource Consumption**: Prometheus can be resource-intensive, especially at scale. We'll need to monitor and optimize resource usage to prevent performance bottlenecks.
+- **Missed Advantages of VictoriaMetrics**: By not adopting VictoriaMetrics, we may miss out on its superior performance in high-cardinality and long-term storage scenarios.
+
+**Risks and Mitigation**:
+
+- **Complexity Management**: To prevent configuration sprawl, we'll establish best practices for managing CRDs and use tools for configuration management.
+- **Training and Documentation**: We'll invest in training sessions and create documentation to bring the team up to speed with the Prometheus Operator.
+- **Performance Monitoring**: Regular audits and performance testing will be conducted to ensure the monitoring system remains efficient as we scale.
+
+---
+
+**References**:
+
+- [Prometheus vs. VictoriaMetrics](https://last9.io/blog/prometheus-vs-victoriametrics/)
+- [Prometheus 3 Beta Announcement](https://www.prometheus.io/blog/2024/09/11/prometheus-3-beta)