DevOpsArk + Prometheus
Read existing metrics, or replace the collector
Vendor: CNCF
What does the DevOpsArk Prometheus integration do?
The DevOpsArk Prometheus integration reads metrics from existing Prometheus servers, aggregates them across clusters, retains them beyond local retention and adds behavioural baselines and cross-signal correlation.
What the Prometheus integration provides
- Query existing Prometheus servers across clusters through one interface.
- Aggregate and retain metrics beyond local Prometheus retention.
- Import Prometheus alerting rules and evaluate them with correlation and suppression.
- Add behavioural baselines so alerting no longer depends only on static thresholds.
- Correlate Prometheus metrics with logs, traces and deployment events.
- Ingest Prometheus exposition format directly where you prefer not to run Prometheus.
Exactly what is read, and what can be written
Read and write are separate grants. Nothing in the write column happens unless you explicitly allow it.
- Time series via the Prometheus HTTP API
- Recording and alerting rule definitions
- Target and scrape health
- Exposition-format endpoints scraped directly
- No writes to Prometheus. Rule changes remain in your own configuration management.
Connecting Prometheus
- 1Point at the server
Provide the Prometheus endpoint, in-cluster or external, with credentials if required.
- 2Map labels
Map Prometheus labels to the DevOpsArk service model so views assemble automatically.
- 3Import rules
Bring existing alerting rules across and add correlation and suppression.
- 4Set retention
Choose how long DevOpsArk retains the aggregated series beyond local retention.
Which modules use Prometheus
Prometheus integration: frequently asked questions
It can, but it does not need to. Many teams keep Prometheus as the collector and use DevOpsArk for cross-cluster aggregation, long-term retention, baselining and correlation with logs, traces and deployments.
Yes. Prometheus servers are queried through their own API, and DevOpsArk views are built on the same series.
Correlation of related alerts into single incidents, suppression during maintenance and rollouts, ownership-based routing and behavioural baselines so rules do not depend entirely on static thresholds.
Yes. Endpoints in Prometheus exposition format can be scraped directly, and OpenTelemetry metrics can be ingested over OTLP.
Integrations that commonly go with this one
Connect Prometheus and see your own data
Read-only connection takes minutes. Most demos start with the customer estate rather than a sample one.